The match between experienced difficulties in ...

89

Technology and Disability 22 (2010) 89–98 DOI 10.3233/TAD-2010-0294 IOS Press

The match between experienced difficulties in everyday activities after stroke and assistive technology for cognitive support Eva Lindqvista,b,∗ and Lena Borella a b

Karolinska Institutet, NVS, Department of Occupational Therapy, Fack 23200, 141 83 Huddinge, Sweden The Municipality of Huddinge, Sweden

Abstract. There is growing interest in computer-based assistive technology for people who experience difficulties related to impaired cognition. More knowledge is, however, needed regarding how this technology can provide support in the everyday activities in which people with cognitive impairment experience difficulties. The aim of this study was to identify in which everyday activities a specific type of computer-based and modifiable assistive technology could provide adequate support to persons who experienced difficulties related to cognition after a stroke. Interviews were conducted with six participants who had experienced a stroke, to identify difficulties related to impaired cognition in everyday activities. The difficulties identified were matched with the assistive technology, using a new tool developed for this study; The Assistive Technology support process. The assistive technology used in the study was judged to be able to initiate the performance of a specific task and to inform about upcoming events. With the use of sensors placed in the home, the support provided related to the completion of an already initiated task or to reminders required in a specific location or after specific actions. This type of support could be used more frequently in the near future and consequently it is necessary to establish what the assistive technology available can and cannot provide as well as what decisions to make before implementation. The Assistive Technology support process could be useful in retrieving this type of information. More research into the implementation of computer-based assistive technology is required.

1. Introduction Within the research field of Assistive Technology (AT) for people with difficulties related to impaired cognition, there has been growing interest for computer-based AT that enables modifications based on the needs of the individual. Such modifications include configurations and information received from sensors in the environment [3]. The aim of this relatively new type of AT is to compensate for a person’s cognitive limitations in the performance of everyday activities. The technology varies in complexity and design and is constantly evolving.

∗ Corresponding

author. E-mail: [email protected].

Studies relating to AT that receives information from sensors in the environment have often been conducted in laboratories or specially designed laboratory flats. In this study the objective was to investigate what possibilities there were for computer-based AT already on the market that uses sensors to support individuals’ experienced cognitive difficulties in everyday activities after a stroke. The assumption was that investigating the possibilities of one of many AT systems would provide knowledge that could also be applied in other systems. In this paper, cognitive difficulties refer to experienced disturbances in the performance of everyday activities which can be seen as related to cognition such as the selection, integration and organisation of information, memory, learning and communication [17]. More than 1.1 million people in the EU countries (and also including Switzerland, Norway and Ice-

ISSN 1055-4181/10/$27.50  2010 – IOS Press and the authors. All rights reserved

90

E. Lindqvist and L. Borell / The Assistive Technology support process Table 1 Socio-demographics of the participants Male/Female Age Living Together/Alone Years after stroke MMSE score at incl. (max 30) Former occupation

P1 M 77 Together 1 27 engineer

P2 M 87 Alone 1 25 clerk

land) [28] experience a stroke every year and between 35% [21] and 72% [16] of those who have experienced a stroke have difficulties related to impaired cognition one year after the stroke. Frequent cognitive problems relate, for example, to the ability to solve problems, memorise and organise, and the most common symptoms are attention deficits, short-term memory problems and aphasia [16]. Persons who experience cognitive impairments after a stroke are often cared for by the local community or by the family and this group of older adults is increasing [28]. It was therefore seen as important in this study to learn more from their experiences. Few examples of computer-based AT for cognition have been found in which sensors are applied in a real home setting. However, in one study by Boman et al. [1], results show that people who had experienced a stroke increased their ability to perform tasks with the help of electronic sensors-equipped AT in the home. In another study [7], and according to the participants, reminders and alarms provided a safe and secure environment when staying in a training flat with sensorequipped AT. Instead of being dependent on their own impaired capacity, the participants in this study relied on the AT support. According to Scherer [24,26], it is important that the design of the AT support is based on the users’ needs and wishes. The AT in the present study, the Tentaculus System (TS) [27], was chosen since the TS was judged, to be able to meet people’s needs and wishes in their everyday activities by applying individual configurations. In this study, everyday activities refer to the activities that a person performs regularly in daily life, like getting dressed in the morning or cooking. Series of tasks [5] constitute the everyday activities. More and comprehensive knowledge is required about the match between what people with impaired cognition experience as difficult in performing daily life activities and what present technology can offer. In this study, a first step was taken in examining these questions. The specific aim of this study was to identify in which everyday activities a specific type of computer-based and modifiable assistive technology could provide adequate support to persons who experienced difficulties related to cognition after a stroke.

P3 F 76 Together 10 27 veterinary

P4 M 77 Together 5 28 engineer

P5 F 69 Together 2 27 nurse

P6 M 76 Together 1 26 electrician

2. Material and methods 2.1. Design Data for this explorative and descriptive study were collected through interviews [15] based on two assessment instruments in the context of an intervention project conducted by researchers at Karolinska Institutet and the Royal Institute of Technology KTH. A number of the participants in the project had had a stroke and they were included in this study.The data were analysed using content analysis [11]. 2.2. The sample Criteria for inclusion in the study were that the person had had a stroke at least one year ago, and had afterwards experienced some kind of difficulty in everyday activities related to cognition. The participants were to be over 65 years of age. It was necessary that all participants and their significant other were capable of being interviewed in Swedish, and that they agreed to the possibility of having AT installed in their home. The participants were recruited by health professionals and via a local stroke association in the Stockholm area of Sweden. The sample can be seen as a convenience sample. Six persons (four men, two women) ranging between 69 and 87 years of age participated in the study (Table 1). One participant lived alone and the others together with a partner. The participants had experienced a stroke between one and ten years ago (median 1.5 years ago), and had faced difficulties in everyday activities related to impaired cognition. Their Mini Mental State Examination (MMSE) [9] mean score was 26.7, ranging between 25 and 28 p/30 p. A suggested cut off score for cognitive impairments after stroke is 27p [2]. Twelve persons who were originally recruited were not included in the study. The reasons were that these persons did not experience problems related to cognition, or conversely, that they were severely affected by their stroke and could therefore not perceive any benefit of the support in their current situation.

E. Lindqvist and L. Borell / The Assistive Technology support process

2.3. The material The AT referred to in this study is the Tentaculus System, TS [27], and consists of a Windows-based laptop equipped with software for individual configurations and external hardware for radio communication between the computer and, for example, loud speakers, different sensors and buttons for interaction with the user. A small cabinet with laptop and hardware for radio communication is installed in the home of the user. The system is connected to the Internet via a mobile modem to enable remote service and configurations. According to the manufacturers, the TS can maintain daily routines, give reminders, support the performance of specific tasks and support the surveillance of the home [27]. 2.4. Data collection Interviews were conducted in the participants’ homes on two occasions in order to identify what difficulties the participants experienced in their everyday activities. The participants were given the opportunity to decide whether a significant other should participate or not. In four of the six cases a significant other was present during the interviews in the role of an additional informant. When the significant other participated in the interview, the researcher documented both persons’ statements and experiences, keeping the statements separated from each other. The interviews were audio-recorded and notes were taken to capture the conversation outside the interviews and also to document the researcher’s reflections around the interview situation. In order to acquire information about the participants’ difficulties in their everyday activities, interviews [15] based on two assessment instruments were conducted; OARS, the Older Americans Resources and Services [8], and CAPM, the Comprehensive Assessment of Prospective Memory [23]. One part in the section A of the OARS instrument, called OARS Multidimensional Functional Assessment Questionnaire (OMFAQ) was used as a basis for the interview relating to everyday activities. Empirical evidence supports the internal construct validity and predictive validity of OMFAQ [12]. In the interview about difficulties related to cognition in everyday activities, the CAPM [23], was applied. Test-retest reliability and internal consistency for the CAPM are within acceptable ranges [4]. Initially, the questions from the questionnaire were asked, thereafter complementary questions derived from the participants’ answers were asked.

91

Data were also collected with the aim to develop a tool applicable in matching the specific type of assistive technology (TS) and the need of support. For that purpose data were collected from the material documented during the implementation of TS in the intervention project (as described in Design) and in which this study was included. The material collected was gathered from every stage of the implementation (i.e. during the identification of difficulties and decisions about what activity to support and the design of the application) and consisted of the recorded interviews, field notes from the communication with the participants and the significant others as well as documentation by the research team relating to possible solutions. Also the communication, that took place before the installation, between the research team and the providers of the technology (Tentaculus Independent Living AB), regarding the planned installations, was included in the data and comprised of, for example, discussions and questions in the form of e-mails or documented telephone calls. 2.5. Data analysis The analysis for this study was conducted in four steps. In the first step, the interview data were analysed to identify difficulties in the everyday activities due to cognitive difficulties. In the second step, a tool was developed to enable an assessment of the match between the identified difficulties and the selected assistive technology, TS. In the third step the difficulties in everyday activities related to cognition identified in the analysis in step one were matched to the TS support by using the tool developed in step two. In the fourth step, the categorisation took place. In the first step the audio files were transcribed verbatim. The interviews and the notes were carefully read in order to identify difficulties related to cognition when engaging in everyday activities. Next a description of each difficulty experienced by the participant, based on all available data, was transcribed, summarised in a few sentences as close as possible to the participants’ own wording and then transformed into lists of difficulties. The lists of the participants’ difficulties were thereafter compiled and the difficulties that were found to be duplicates were removed. Brief explanatory descriptions of the difficulties were recorded. In the second step of the analysis, a tool for assessing the match between the identified difficult tasks and the possible required support by TS was developed since no tool designed to perform such a match had been found. A content analysis [11] of material collected during the

92

E. Lindqvist and L. Borell / The Assistive Technology support process Table 2 The four steps of The AT support process with related questions Receive information What kind of calendar information is required? (weekday, date) What personal calendar information is required? (birthdays, holidays) What events do they wish to be informed about? (appointments, activities) Are the events recurrent/irregular/single? From what area/object should information be received? What type of sensor is suitable for receiving the desired information? (e g electric sensors, magnetic sensors) What restrictions for receiving sensor information should be set? (e.g. detection area) At what hours should the sensors be active? Deliver information At what time should the chosen information (recurrent/irregular/single) be presented? Where in the building should the user be alerted? In what form should the information be delivered? (e g voice, sound, light) Is repetition desired? Should the repetition of information be active or passive? At what interval should the information be repeated? For how long should the repetition last? How should information be delivered continuously, if desired? After sensor being triggered: When should the user be alerted? ( directly, after a specified amount of time, or after predetermined sensors have acted in a specific order) Confirm the task Is it desired and/or possible to choose passive confirmation? What sensors should be triggered to indicate that the task is performed? What actions should be conducted if the chosen sensor does not get confirmation? What actions should be conducted if the chosen sensor gets confirmation at an inappropriate time? Additional actions What additional actions are appropriate? What events should activate an additional action? After what length of time should the action be conducted? Should it be possible for the user/significant other to turn off the support? How? What events should activate the automatic stop and restart?

implementation of TS was performed. The aim was to identify the information and functionality needed for the design of the individual support, support in which the information inserted is tailored to the individual’s preferences and habits. Thereafter data were labeled and grouped according to the features of the data. This process resulted in four steps and they were defined as follows: 1) Receive information concerns how the AT receives information to act upon. The instructions in this step result in as accurate information as possible to reduce unnecessary alerts while at the same time not fail to give an expected reminder. 2) Deliver information concerns how the AT alerts the user. The instructions in this step should be formulated in such a manner that the desired message should be as accurate as possible with respect to time, place, task and person. 3) Confirm information concerns how the AT receives confirmation that the user has performed the expected task. The instructions in this step should be formulated in order to reassure the user and significant others as to whether a task has been performed or not. 4) Additional actions concerns whether

the performance of any additional actions, with the exception of alerting, is adequate to support the user. Also included in this step is the possibility of stopping the support manually or automatically. The four steps made up the AT support process which can be seen as a generic process for support from AT of this type. The data collected during each step in the process was used to develop the AT support process questionnaire which included questions judged as requiring answers (Table 2) when matching the AT with a specific task. In the third step of the analysis, the matching process was conducted. Firstly, for every task the participants reported as being difficult the prerequisites for the provision of support by the AT were identified by the first author, an occupational therapist, who had taken part in the implementations of the TS within the project and was consequently familiar with the product. In this step, the AT support process questionnaire was used (Table 2). Thereafter, the prerequisites for the tasks were matched with the specific AT in this study: TS. In doing this the question asked was whether TS had the


potential to perform the necessary actions in each step and the answers were: “yes”, “no” and “maybe”. Finally, an assessment was made using the result from the previously conducted match regarding to what extent the TS had the potential to support the tasks reported as being difficult. The assessment was based on descriptions of the first two steps in the AT support process; to receive and deliver information, since without receiving information, the support would not react and without delivering information the user would not get any information to react upon. When TS could not perform the actions needed in these two steps, it was assessed that the support was insufficient. If passive confirmation or additional actions were not possible to obtain, it was judged that the support might still be beneficial. To validate the findings in the above three steps, a consensus development method influenced by the Delphi method [20] was implemented. Questionnaires were sent to two assessors, two occupational therapists, experienced in working with TS in research. The question asked for each activity was whether the task or activity could be supported by TS and was to be answered with a “yes” or a “no”. There was also space available for making comments. The author discussed the responses individually with the assessors in a structured manner to achieve an understanding of the underlying reasoning. This resulted in alternative solutions in three instances. In the fourth and final step of the analysis, the categorisation, it was recognised that the step relating to how the TS received information to react upon was the step in which the instructions for the support differed the most. That step was also recognised as being the most significant step when assessing whether it was possible to obtain support or not. Consequently, the tasks analysed were divided into how TS received the information to be reacted upon. The five types of information were 1) calendar information only, 2) inserted individual calendar information, recurrent, 3) inserted individual calendar information, irregular or single, 4) from electronic sensors only and 5) from various sensors. Recurrent and irregular/single reminders were divided into two groups since they differ in the manner of how the administrator inserts the information. The regularly recurrent reminders are inserted once for a number of reminders, but irregular or single reminders are inserted separately. In Table 3, the TS version of 2007 is presented according to the AT support process. The information is grouped in line with the above identified types of reception.

93

For a number of tasks, it was assessed that TS could not provide sufficient support. These tasks were divided into groups according to their similarities regarding the reason why the support was judged as not being possible to provide.

3. Results 3.1. Tasks assessed as possible to support It was assessed that the TS was able to support 29 of the 65 reported tasks in which the participants had experienced difficulties related to cognition. In Table 4, these tasks are presented in sections according to how the TS was expected to receive the required information as described in the AT support process. Below, the results are described briefly and in accordance with the table structure. The TS receives information from two types of sources, from the calendar inside the computer (group 1, 2, 3 in Table 4) and from sensors in the environment (group 4, 5 in Table 4). It was assessed that the calendar information was useful in two cases; regarding information about day and date and regarding different types of prospective reminders. It was determined that the reported desire to have information about day and date (group 1, Table 4) was supportable by a synthetic voice, which could give the user the calendar information he or she wanted. By inserting personalised information in the calendar in the computer, it was determined as possible to design recurrent, irregular and single reminders (group 2, 3, Table 4). The tasks were either related to initiating the performance of a task or to informing the person about coming events. Examples of tasks that were judged to be in this category were remembering to water the plants or being informed about plans for the day. Tasks reported to be regarded as suitable for frequent reminders were those that had to be performed regularly like taking medicine or tasks that were desired to be performed regularly, like taking a walk. When information from sensors in the home were regarded as suitable for support of the reported difficulties (group 4, 5, Table 4), these tasks were often related to the support of the completion of an already initiated task, for example to remember to hang the laundry left in the washing machine after washing, turn off the cooker (group 4, Table 4) or to lock the door (group 5, Table 4).

94

E. Lindqvist and L. Borell / The Assistive Technology support process Table 3 A description of TS, version of 2007, according to the AT support process support

Receive information Deliver information 1) Tasks in which calendar information only is needed Synthetic voice message at a specific time, or on General calendar information, peruser’s request when pressing a button sonalised by decisions regarding Repetition, active or passive limits of the information Constant information not possible 2) Tasks in which personalised calendar information (recurrent events) is needed Information about recurrent events Synthetic or recorded voice message regularly, inserted in the calendar by adminisFor alerting, combined with light or sound Text message to mobile phone trator Repetition at predetermined points in time or Repeated according to user’s desires when the user presses a button Constant information with signs/word on a lit-up sign

3) Tasks in which personalised calendar information (irregular/single events) is needed Information about single/irregular Synthetic or recorded voice message at predeterevents inserted in the calendar by mined single occasions For alerting, combined with light or sound Text administrator message to mobile phone Repeated according to user’s desires Repetition at predetermined points in time or when the user presses a button 4) Tasks in which information from electronic sensors only is needed Information received from sensors Voice, sound, light or sign message immediately placed between the electric plug and or after a preprogrammed amount of time after the socket, when necessary with the sensor is triggered delay due to irregular electricity Repeated in predetermined intervals consumption of certain items (e g For alerting, combined with light or sound cooker, coffee machine, washing Text message to mobile phone machine) Constant information with signs/word on a lit-up sign, indicating an on-going action 5) Tasks in which information from various sensors is needed Voice or sign message immediately or after a preInformation from electric, pressure programmed amount of time after the sensors are or magnetic sensors (mounted on triggered or after that predetermined sensors have doors, windows and inside locks), reacted in a specific order. motion detectors and IR sensors in Repeated in predetermined intervals or when senthe home environment. sors are or are not triggered according to the plan. For alerting, combined with light or sound. Text message to mobile phone Constant information with signs/word on a lit-up sign indicating an on-going action

Other tasks considered to be in need of sensors for support were reminders associated with a specific place or to a specific action. For example, when sensors register movements in the hallway at a specific time, the predetermined assumption is that the user will go out and is in need of a reminder about, for example, taking their walking stick with them. Another example is when the system (TS) has registered that the cooker has been used and subsequently a sensor by the kitchen door is triggered. In this case, an oral message can be delivered that the person might need to tidy up before leaving the kitchen.

Confirm task passively

Add. actions

None

None

Magnetic sensors placed on relevant doors (medicine cabinet, freezer, front door) Motion detector close to relevant items Electric sensor on electronics, only active during a limited period of time

None

None

None

Electric sensor by e g the cooker or by the coffee machine. Magnetic sensor by washing machine door

Turn off electricity in line with predetermined terms

Magnetic sensors by kitchen cupboard, doors and door lock. IR sensor by basin or toilet

None

Those examples represent tasks in which it was assessed that the support would very probably deliver a large number of unwanted reminders (that is, reminders for the wrong person or for the right person at the wrong time). To reduce the amount of unwanted reminders it was established that time limits and a minimised detection area were probably needed. 3.2. Tasks a ssessed as not possible to support For a number of tasks, it was assessed that it was not possible for TS to provide support (see Table 5).


95

Table 4 Tasks assessed as possible to support by TS, version of 2007 1) Tasks in which calendar information only is needed Be informed day and date about: 2) Tasks in which personalised calendar information (recurrent events) is needed Remember look at the calendar, Take initiative regularly to: take the medicine, regularly to: eat, change clothes, Be informed put the rubbish bin out about: for pick-up, water the plants, check the phone receiver. 3) Tasks in which personalised calendar information (irregular/single events) is needed Remember to: make a planned phone call, Be informed keep an appointment about:

use exercise bike, practise speech therapy, go for a walk, contact relatives. birthdays at the right time.

plans for the day

4) Tasks in which information from electronic sensors only is needed Remember to: turn off the cooker, turn off the coffee machine, hang the wet laundry left in the washing machine. 5) Tasks in which information from various sensors is needed Remember to: bring the wallet/walking stick when leaving home, tie shoes, flush the toilet, turn off the water tap, lock the door, check that the door is locked, close the doors of the kitchen cupboards, tidy up after cooking

One reason was the difficulty in receiving adequate information to give useful support. For example, since the TS could not provide sensors for small objects, the support requirement relating to looking for and finding ones glasses or mobile phone could not be supported by this system (group 1, Table 5). Neither was it judged that the TS could provide support when handling electronics (group 2, Table 5), like phones and computers. The users for example often experienced a need to comprehend information on screens or displays and to complete a series of steps according to the information provided. When the user was in need of reminders based on recently received information (group 3, Table 5), like the reported difficulty of handling changes in plans, the TS was judged as not being supportive. The TS system was not designed to enable the user or significant other to insert reminders in the system and this service could not be provided by the administrator at short notice. This limitation also meant that it was judged that the TS was not able to support the user in the event of them wanting to pass on a message to someone. A number of the reported difficulties were activities, or

series of tasks (group 4, Table 5). The performance of these activities was dependent on the individual’s own intentions and decisions within a specific context at a specific point in time, e.g. making dinner and playing cards. Activities that had these characteristics were judged as not supportable by the TS. Nevertheless, support was assessed to be possible in some tasks involved in for example the activity cooking, such as getting a reminder to turn off the cooker. A prerequisite is that the user can choose the right item for the task and also manipulate the item adequately. Regarding communication (group 5, Table 5), the reported difficulties were foremostly related to conversation with other people, but also to being able to remember and use information. Neither of these tasks was judged as supportable by the TS, since as described earlier, the TS software was not designed to allow the user to insert any type of information. 4. Discussion The findings in this study show that computer-based AT that can be modified with respect to individual re-

96

E. Lindqvist and L. Borell / The Assistive Technology support process Table 5 The tasks assessed as not possible to support by TS 1) Tasks related to tracking items Find glasses, Find mobile phone, Find items in the kitchen cupboard. 2) Tasks related to handling technology Pay bills on the internet, Buy and sell shares on the Internet, Look for information on the Internet, Order books from library via the Internet, Remember number from phone book when dialling, Use the mobile phone. 3) Tasks in which insertion of reminders or information by user or significant other is required Book appointments, Handle changes in the daily routine, Handle changes in medical prescription, Remember to buy an item at the shops, Remember to pass on a message, Remember written information, Retell something read in the paper, Remember characters when reading a novel, Handle calculation when selling items, Practice mental calculation. 4) Tasks influenced by the person’s decisions at a specific moment in a specific context Initiate, be focused during, and finish activities in general, Remember all clothes items when dressing, Turn the clothes the right way out, Make dinner, Remember all ingredients when cooking, Make decision regarding the household together with spouse, Give instructions to spouse regarding household activities, Play cards. 5) Tasks related to communication Talk native language, Alternate native language and present language, Have a conversation in a small group, Remember to mention a point during conversation, Avoid repeated wordings, Remember names of familiar people, Draw maps. 6) Tasks performed outside home Pour antifreeze mixture in the right place, Pay the right amount of money when shopping.

quirements could be considered in the following cases: 1) when the inbuilt calendar is used, the support is most beneficial when prospective reminders are needed on a regular basis in activities that are possible to plan, 2) when sensors are used, the most appropriate support is shown to be when the sensors are triggered by the actual event which the user wants to be informed about, for example, that the washing machine has stopped and the wet laundry should be taken care of. Support from the AT is regarded as being less reliable but still purposeful, when sensors are triggered by actions that signal that an event has taken place or will take place, for example, a reminder given in the hallway to alert the users to take their wallet with them when leaving home. Even if time-limits are set and the detection area is minimised unnecessary or undesired reminders will occur, which could cause confusion or annoyance. However, in some cases it might be more beneficial to get a number of undesired reminders together with the desired reminders than no reminders at all. To minimise the amount of unnecessary reminders the Autominder [22] could be an option. The Autominder has been developed with the aim to increase the efficiency of the support regarding the management of the user’s daily schedule, partly by handling information from monitors which register the actions performed in the home [22]. TS was judged to be unable to provide support in some situations, however there could be other products on the market or AT designed for a specific purpose which could be suitable, as described by Gorman et al. [10]. Examples of these products are locators of

glasses and mobile phones, and portable memory aids. Software for computers or personal digital assistants designed for ease of use when for example inserting reminders or navigating on websites is also available. Computers may also provide possible communication support, for example in the use of e-mail. This type of support would not be beneficial for the users in this study since their difficulties were more related to finding words and remembering to bring up a specific subject when meeting people. There is also a risk that, even if the software is designed for people with cognitive impairment, it might be too complicated to use. As reported, the AT was not judged to be supportive in activities which included series of steps that were influenced by the user’s intentions and the context in combination with many complex manual manipulations, e.g. making dinner or getting dressed. These activities have been described in previous studies [6,19,30]. One such reported example of problematic activities is turning clothes the right way out when getting dressed. In cases such as this, the AT can remind the user to get dressed, but cannot provide guidance on how to put on their clothes. Studies have however been conducted in which the user has been guided by supportive reminders in activities including just a few steps [10,18]. Sensors would perhaps enhance the accuracy of the support in cases in which the user is guided in the activity by the AT. Nevertheless, studies that have monitored people performing everyday activities required a large amount of sensors of different types to enable the registration of the person’s doings and to be able to give relevant re-


minders in more complex situations [19]. There is a risk that even if a person’s habits and cognitive problems are identified, it might be difficult to understand the person’s intentions at a given moment and to thereby provide suitable support. A tool for the match between the TS and the need of support was developed; the AT support process, which aimed to identify how the AT was able to give support in specific tasks. In order to receive information for each step in the AT support process, 26 questions were developed. Only a few of these questions address the technical issues. The majority of the questions should be answered by the users and significant others for the personalisation of the support. With respect to the clinical implications, it is important to bear in mind that valuable information is, to a great extent, obtained from the users’ and the significant others’ experiences and their abilities, desires, needs, habits, values and contexts when highly technical AT is implemented. It is important to be aware of these aspects in these, as in all client-centered interventions related to everyday activities [14]. Previous studies have also demonstrated the importance of these aspects when planning an AT intervention [6,25,29]. When it has been established that a type of modifiable AT support might be suitable, the AT support process presented in this study could be useful. By asking the questions outlined here, responses can be obtained on how the support for the specific user should be designed and also whether it is at all possible to implement this type of AT solution. Assessments should be made prior to installation as to whether an AT solution is likely to give the intended support or not [6]. One key reason for this is that it could save time and effort for all involved and is in line with the view presented by Chan [3]; that inappropriate technology should be identified and excluded as soon as possible. Furthermore, when support appears to be appropriate, and personalised support has been tailored, it could still be the case that the AT proves to be unsuitable after all after the installation has taken place. One reason might be that even if the support works as planned, the user does not perform the expected task, perhaps due to limited motivation or understanding [13]. Furthermore, even if the task is performed and the user’s goal is achieved, the support might not have the positive effect on everyday activities that the user expected or it could be experienced as awkward when finally in place. Another risk is that problems with the support such as electrical short-circuiting, computer failure [1] or delayed text messages from the phone operator are

97

perceived as stressful to the user. This underlines the importance of following up after the installation of AT and learning about how the AT is experienced by the user. It is important to remember that the reason for applying the AT is to empower the user [29] and to bring a “meaningful change” to the user [13]. Many studies, including this one, have in a variety of ways shown that the involvement of the users and significant others is crucial if this is to be achieved [3,24,29]. Moreover, when the support is appreciated there is reason to follow up and make new assessments as to whether the AT can support the user’s new or altered requirements since all of the individual requirements are not detected at first and they can also change over time [1,6]. It was deemed necessary to develop a specific tool for the match procedure since no tools were found for this type of match. This tool, the AT support process, which highlights how the judgments included in the match were made, became of significant importance of the study, even though the development of the tool was not aimed to be in focus. The reason was that the question why the AT was judged to match a task appeared to be equally interesting as the question what tasks the AT was judged to match. Consequently, great effort has been put into making the steps in the match procedure transparent and the match was conducted by initiated judgments by experts available. Additionally, two experts were invited to make judgments to ensure efficient validation. To conclude, the result of this study demonstrates that this type of support might have the potential to support people with cognitive impairments in the performance of their everyday activities. These AT solutions could be used more frequently in the near future than they are today. In that case, it is crucial that health professionals have knowledge about what the specific AT can and cannot provide, and also about what aspects it will be necessary to take into consideration to ensure that the AT is appropriately adjusted to the specific user and environment. Hopefully the development of the AT support process in this study, can provide a first step towards a guide for these implementations. However, more research about how to match the person with this type of computer-based AT is needed.

Acknowledgements We would like to thank the participants, who kindly shared their thoughts and opinions with us. We also

98


would like to thank Elisabeth Lagerkrans and Ingalill Boman, colleagues who took of their time to discuss the support as a part of the analysis, and Tentaculus Independent Living AB that provided equipment and services. No other relations or dependences existed between the research team and the providers. This project was funded by KTH Royal Institute of Technology and Stockholm County Council via CTV and by the municipality of Huddinge with support from the Swedish Brain Power network.

[12]

[13]

[14]

[15] [16]

References [17] [1]

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

I.L. Boman, A. Bartfai, L. Borell, K. Tham and H. Hemmingsson, Support in everyday activities with a home-based electronic memory aid for persons with memory deficits, Disability and Rehabilitation: Assistive Technology, in press. A. Bour, S. Rasquin, A. Boreas, M. Limburg and F. Verhey, How predictive is the MMSE for cognitive performance after stroke, J Neurol 257 (2010), 630–637. M. Chan, D. Estève, C. Escriba and E. Campo, A review of smart homes- present state and future challenges, Computer Methods and Programs in Biomedicine 91 (2008), 55–81. L.T. Chau, J.B. Lee, J. Fleming, N. Roche and D. Shum, Reliability and normative data for the Comprehensive Assessment of Prospective Memory (CAPM), Neuropsychological Rehabilitation 17 (2007), 707–722. C. Christiansen and E. Townsend, Introduction to Occupation, the art and Science of Living, (2nd edition), P102-3 Pearson Education, Upper Saddle River, New Jersey, 2010. G. Dewsbury, B. Taylor and M. Edge. The process of designing appropriate smart homes: including the user in the design, 1st Equator IRC Workshop on Ubiquitous Computing in Domestic Environments, The school of computer science and information technology, the University of Nottingham, 13–14th September 2001. A. Eriksson, G. Karlsson, M. Söderström and K. Tham, A triaining apartment with Electronic Aids to Daily Living: Lived experiences of persons with brain damage, The American Journal of Occupational Therapy 58 (2004), 261–271. G.G. Fillenbaum. Multidimensional Functional Assessment of Older adults: The Duke Older Americans Resources and Services Procedures, Hilldale, NJ: Lawrence Erlbaum Publishers 1998. M.F. Folstein, S.E. Folstein and P.R. McHugh, “Mini-Mental State” A Practical Method for Grading the Cognitive State of Patients for the Clinician, Journal of Psychiatric Research 12 (1975), 189–198. P. Gorman, R. Dayle, C. Hood and L. Rumrell, Effectiveness of the ISAAC cognitive prosthetic system for improving rehabilitation outcomes with neurofunctional impairment, Neurorehabilitation 18 (2003), 57–67. U.H. Graneheim and B. Lundman, Qualitative content analysis in nursing research: concepts, procedures and measures to achieve Trustworthiness, Nurse Education Today 24 (2004), 105–112.

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27] [28]

[29]

[30]

K.L. Haywood and A.M. Garratt, R. Fitzpatrick. Older people specific health status and quality of life: a structured review of self-assessed instruments, Journal of Evaluation in Clinical Practice 11 (2005), 315–327. N. Kapur, E.L. Glisky and B.A. Wilson, Technological memory aids for people with memory deficits, Neuropsychological Rehabilitation 12 (2004), 41–60. G. Kielhofner, Model of Human Occupation: Theory and Application (3rd edition), Lippincott, Williams & Wilkins, 2002. S. Kvale, Interviews, An Introduction to Qualitative Research, Sage publication, 1996. M. Lesniak, T. Bak, W. Czepiel, J. Seniow and A. Czlonkowska, Frequency and prognostic value of cognitive disorders in stroke patients, Dement Geriatr Cogn Disord 26 (2008), 356– 363. M.L. Lezak, D.B. Howieson and D.W. Loring, Neuropsychological Assessment. Oxford, University Press, 2004. E.F. LoPresti, R.C. Simpson, N. Kirsch, D. Schreckenghost and S. Hayashi, Distributed cognitive aid with scheduling and interactive task guidance, J of Rehabilitation Research & Development 45 (2008), 505–522. A. Mihailidis, F. Fernie and W. Cleghorn, The development of a computerized cueing device to help people with dementia to be more independent, Technology and Disability 13 (2000), 23–40. M. Murphy, N.A. Black and D.L. Lamping, Consensus development methods, and their use in clinical guideline development, Health Technology Assessment 2 (1998), 1–276. M. Patel, C. Coshall, A.G. Rudd and C. Wolfe, Natural history of cognitive impairment after stroke and factors associated with its recovery, Clinical Rehabilitation 17 (2003), 158–166. M. Pollack, Intelligent Technology for an aging population, the use of AI to assist elders with cognitive impairment, AI Magazine 26 (2005), 9–24. N.L. Roche, J.M. Fleming and D.H.K. Shum, Self-awareness of prospective memory failure in adults with traumatic brain injury, Brain Injury 16 (2002), 931–945. M.J. Scherer and G. Craddock, Matching person & technology (mpt) assessment process, Technology and Disability 14 (2002), 125–131. M.J. Scherer, The change in emphasis from people to person: introduction to the special issue on Assistive Technology, Disability and Rehabilitation 24 (2002), 1–4. M.J. Scherer, Assessing the benefits of using assistive technologies and other supports for thinking, remembering and learning, Disability and Rehabilitation 27 (2005), 731–739. Tentaculus Independent Living AB, www.tentaculus.com. T. Truelsen, B. Piechowski-Józiak, R. Bonita, C. Mathers, J. Bogousslavsky and G. Boysen, Stroke Incidence and Prevalence in Europe: A Review of Available Data 13 (2006), 581– 598. S. Wey, One size does not fit all: Person-centred approaches to the use of assistive technology, in: Perspectives on Rehabilitation and Dementia, M. Marshal, ed., London, GBR: Jessica Kingsley publishers, 2004, pp. 202–208. J.P. Wherton and A.F. Monk, Technological opportunities for supporting people with dementia who are living at home, Int J Human-Computer Studies 66 (2008), 571–586.

147


Barriers to the adoption of cell phones for older people with impairments in the USA: Results from an expert review and field study Robert Pedlowa,∗, Devva Kasnitzb and Russell Shuttleworthc a

Independent Disability Studies Scholar, Canberra, Australia University of California, Berkeley, CA, USA c Faculty of Health Sciences, University of Sydney, Sydney, NSW, Australia b

Abstract. This qualitative exploratory study examines the barriers to the adoption of off the shelf cellular telephones for older people with vision, hearing, or dexterity impairments. We include the full range of factors that influence phone acquisition, use, and retention. We explore the feasibility of providing individual support to overcome learning barriers. The first stage of the study is a review of the sales process, purchase experience, and handset features available for cell phones in the Berkeley and Oakland areas of California in mid 2007. As the industry is organized, physical accessibility cannot be separated from programmatic access and business practice. The ergonomic aspects of handset design are mediated by firmware and software that changes rapidly and varies by carrier. The US sales model for cell phones, where purchasers select a pricing plan or tariff and then a handset from a limited group, does not accommodate the needs of seniors with impairments who need low cost plans but handsets with specific groups of features, most of which are generally available, but not necessarily available on any one phone and almost never on entry level phones. It is difficult for consumers to obtain information prior to purchase about handset accessibility. Features that support access are often difficult to find in menus and difficult for seniors to use. In the second stage of the study we gave a group of twelve older people with impairments, ranging from 66–90 years of age, a cell phone, airtime, and individual support for a one-month period. The results show that older people with impairments are deterred from cell phone use as much by the confusing structure of the industry as by the lack of certain handset features. They too often sacrifice access for price. They can be supported to learn to use cell phones but further work is needed to determine the specific kinds of support needed and how this can best be delivered. Keywords: Cellular phone, older people, disability, impairment

Every day, millions of people around the world who have a disability, are faced with frustrating – even impossible – situations. ITU believes that these people should enjoy the same services and opportunities in life as everyone else.” The International Telecommunications Union [1]. 1. Introduction The cell phone is emerging as a near universal plat∗ Address for correspondence: Dr. Robert Pedlow, 47 Elkedra Avenue, Hawker, ACT, Canberra, 2614, Australia. Tel.: +61 2 6223 2273; E-mail: robert [email protected].

form for communication all over the world. Counting each individual phone card purchased, European data for 2006 suggests that there are more cell phone subscriptions than people [2]. The developing world is skipping landlines altogether. Phones no longer occupy a fixed place and belong to a group. They are an intensely individual, intimate prosthetic. The “telephone table” is an antique bit of material culture. “Hello, where are you?” used to mean “Which extension are you on, upstairs or down?” Wong has called this the process of “de-territorization” of phones. In Hong Kong, where cell phone penetration is over 84% they have changed cultural patterns of “borrowing” land lines among neighbours and at stores [3]. Horst and


148

R. Pedlow et al. / Barriers to the adoption of cell phones for older people with impairments in the USA

Miller writing in 2006 demonstrate how short but frequent use of cell phones among poor Jamaicans actually saves money in the long run by minimizing wasted effort and maximizing cooperative opportunities. Similarly, younger disabled people find cell phones critical to avoiding wasted effort, which usually also means extra cost [4]. This is an economic equation that should also have resonance for older people as well [5]. Cell phones, like other technology-based products, are designed and marketed for young, able, dexterous users who are comfortable with and enjoy using technology as an everyday part of their lives, often occupying a great deal of time, and accounting for a large proportion of their social interaction [6]. For many a phone is far more than a phone. For older people, the experience of functional impairment – such as mobility impairment, vision loss, or hearing loss – can be tremendously frustrating, depressing, and isolating. This can be exacerbated by short term memory loss [42]. Many older people with these impairments, who could potentially derive great benefit from the improved sense of security, freedom of movement, and improved social contact that cell phones can provide, are unable to use currently available products. Most research on technology for older people has focused on hightechnology safety surveillance options that will require substantial investment to implement [8]. Here in this research we have modest goals, by phone accessibility or useability we refer to the basic function of making and receiving calls, accessing voice mail, and using stored phone numbers. The Rehabilitation Engineering Research Center on Wireless Technologies noted that in their research on policy needs “Device incompatibility or poor interoperability was cited as the most important technology issue . . . Second, the number one awareness/access goal . . . was to encourage manufacturers of wireless devices to include persons with disabilities in the review and evaluation of assistive or universally designed products and technologies.” [9] “This year, World Telecommunication and Information Society Day has adopted the theme: “Connecting Persons with Disabilities: ICT Opportunities for All” to address the special requirements of persons with disabilities.” [10] 1.1. Use of cell phones and age and disability While usage of cell phones by older people is increasing it is still lower than that of younger users. A survey by the American Association of Retired Persons reports that wireless telephones have become an

essential part of life for many older Americans, with respondents age 50–64 being almost as likely as those age 18–49 to report having cell phone service and those over age 65 the fastest-growing age group of cell phone users. The primary reasons for cell phone use for those 65 and over is security [11]. Our strong impression is that most seniors are introduced to cellular phones by family members of whom they are often dependent in learning how to use the phone [3]. We could find no US statistics on this point. Also, rates of cell phone usage specifically by older people with functional impairments are largely unavailable, however 30% of Americans aged 75-94 have a cell phone [12] while UK evidence indicates that cell phone usage rates drop to 24% for people aged 75 and over [13]. The 75 and above age group is where the prevalence of impairments is highest. The irony of this is that the very factors that have made cell phones such a ubiquitous product – the improved sense of security, freedom of movement, and improved social contact – would be of particular benefit to this group. It may be argued that market-driven design improvements and/or the aging of people already familiar with cell phones and other technology means that this lower level of usage will be transitional in nature. However, Vanderheiden’s major study found that the needs of people with impairments are not addressed by the market; a phenomenon that tends to be repeated with each new product and technology [14]. Accessibility features are too often added to a device in the after market, but by that time the base technology has changed. For example, one “simple” cell phone for people with cognitive impairments that displayed four to six photographs to touch for dialling was built around a standard Palm PDA and cost over US $800; this was reimbursable through insurance or paid by US government programs, long before that feature started to appear in mass market phones. However, the base Palm technology was outdated by the time the accommodative “fix” was released. The market phone embedded that feature too deeply in other menus to be usable by the people for whom they were originally designed. The same is true of much voice recognition technology for people who have serious hand dexterity issues. Without true voice navigation, voice activated typing is not always accessible. Even the principles of “universal design” leaves gaps. “Universal design is the process of creating products that are usable by as wide a range of people as is commercially possible.” [4, p. 33] Most people acknowledge that “universal” is really only expected to fit 90% of a population.


A founding member of the universal design movement, Ron Mace defined it as “The design of products and environments to be usable by all people, to the greatest extent possible, without the need for adaptation or specialized design.“ He details several principals embedded in universal design: Equitable Use, Flexibility in Use, simple and intuitive, Perceptible Information, Tolerance for Error, Low Physical Effort, Size and Space for Approach and Use [15]. The need for accessibility features of land lines has been accepted by various governmental entities. For example, the state of California supplies phones, free of charge, with no means test, to people with vision, hearing, dexterity, mobility, speech, and cognitive impairments (see www.ddpt.org/ctap/). These programs do not extend to cell phones. From other work, we were aware that only a few modern cell phones are designed for the access needs of older people or of anyone with vision, hearing, or dexterity impairments [16]. Research by Nguyen et al. showed that in a sample of 47 disabled people, most were not using existing phone features that would help them and that standardization is a challenging problem [17, p. 79]. They subsequently conducted a study much like our research with 10 disabled elderly people, who exhibited three levels of severity of impairment between the ages of 14–80 in a three week trial using off the shelf phones. They conclude that “The successful outcomes obtained required knowledge of all of the features on the available phones, a careful assessment of the participant’s needs and abilities, and comprehensive user training, both initially and when required, during the trial” [17, p. 90]. Younger disabled people are organized to impact the after-market industry, if not the overall regulations. For example, a lawsuit from a blind advocate, Bonnie O’Day, was responsible for new regulations that all handsets must be hearing aid compatible [18]. Other advocates and researchers have had a role in developing, testing, and/or selling specialized hardware or software, much of which is only in Beta, difficult to find, very expensive, and/or very specialized. The US National Institute of Disability and Rehabilitation Research funds a large Rehabilitation Engineering Center on Wireless Technologies that purports to have consumer involvement of disabled people [19], and a second Rehabilitation Engineering Research Center on Telecommunications Access, a partnership between the Technology Access Program of Gallaudet University and the Trace Center of the University of Wisconsin – Madison [20]. They do a lot of work on access to emer-

149

gency functions of telecommunications, and deaf/blind, and PDA/TTY convergence, and relay services, including wireless applications. “In the case of cellular telephone technology, legal mandates have been used to provide some level of accessibility on closed device platforms.” [21] Very little of this work directly considers the elderly with impairments. Wong’s study investigated cell phone use in Hong Kong nursing homes and he did consider “user friendliness” in terms of vision, hearing, and dexterity as issues [3]. Although there is a growing body of general literature on the differences between cohorts aging with disability and aging into disability [22], this is rarely considered in design. 1.2. Purpose of the study In this preliminary heuristic and qualitative research we aimed to begin identifying barriers to accessing cell phones in the purchasing process and design of readily available mass-market phone handsets, as well as in service plans and add on products. We also explored the issue of support to enable older people with impairments to become cell phone users. Enabling older people with impairments to use everyday technology is important to their freedom of choice and equal opportunity to choose their own lifestyle. We identified what we thought are compelling reasons why access to cell phones could offer great benefits for this group. Depending on their living arrangements, disabled elderly could particularly benefit by an improved (i) sense of confidence and security when travelling outside the home and (ii) access to social support [23,24]. Cell phones can potentially improve disabled elders sense of confidence and security when travelling by giving them a means of summoning assistance and improve their access to social support by enabling them to communicate with family when they want to during the day. Summoning assistance: Older people with impairments often experience concern being alone at home or travelling outside their home due to health issues or concerns about falling or personal safety. With the significant reduction in numbers of pay phones in public spaces, older people may be concerned about travelling by public transport or driving. Falls among older people are a serious health concern and may lead to injury, admission to residential care, or even premature death e.g. [25,26]. A cell phone on the person, for example, offers obvious advantages over a wireless landline handset or a dedicated “panic button” technology be-

150


cause it can extend out of the home, is less stigmatised, and more flexible. Communication with family: Too many older people are at risk of social isolation. This is exacerbated if they can’t use a telephone due to a vision or hearing problem. People with motor restrictions can experience significant difficulties in using fixed phones due to the difficulty of reaching the phone in time to receive the call, or because fixed phones may be in an inaccessible locations. A cell phone can overcome these difficulties inside the home and during excursions. For elders in residential facilities, a cell phone can help keep them connected to family [3]. The potential of cell phones to meet these needs for older people is likely to be influenced by a range of factors including the match between the older person’s capabilities and the accommodations provided by the device as well as the living arrangements of the older person and their goals and preferences. Those most likely to have help using a phone may be those who live near younger family members. Currently available cell phones do not readily enable access for people with disability of any age, especially the older disabled population. It is reasonable to ask whether accommodations developed by or for younger disabled people work for disabled elders. Some of the basic issues around control, choice, risk, and agency are the same. As mentioned above, disabled people have seen specialized technology such as voice recognition software become democratised and mainstream but lose its sensitivity so that, for example, it can’t handle disabled people with unusual voices. And, the first phones fitted with photographic address books were after market adaptations for people with cognitive impairments who cannot read. While this technology has become standard on most camera-equipped handsets, other features still make them inaccessible to the cognitively impaired. Phones targeted for particular age groups are an emerging concept in cellular handsets. The most common are “senior phones” – that is, phones for older people that have begun to appear in the North American, European, and Japanese markets. They provide a simple handset with a small number of functions. In the US market, several phones of this type (e.g., the Jitterbug [27]) are being marketed as senior phones. New companies are adding innovative competing products all the time. At least four new “senior” phones have been released since the time of this research, none of which appear to have solved most of the issues in our findings [28]. Blogs and websites to assist seniors

choose a phone are also proliferating [29]. Motorola is developing software that recognizes an older person’s voice and would then automatically adjust cell phone handset settings, an authoritarian solution apparently still far from actually being implemented. Worldwide, the major focus of research on technology to support older people in living independently has been to develop specialised products such as embedded sensing devices for home monitoring and emergencies [2]. These products tend to be expensive, and the approach has neglected the needs of older people with impairment for choice, control, and mobility outside their homes. Also there is increasing evidence that the success and acceptance of technology, especially for older people, depends not only on the functionality but also on how the person using the technology feels that they are being treated by others [30]. Instead of developing specialized high-technology solutions that are likely to identify older people who use them as being “sick,” the current heuristic research aims to explore if, when, and how we can support older people with impairments so that they can use everyday technology, or everyday technology with minimal modifications. The concept of providing support to enable older people to use everyday technology is an approach that could potentially be made available relatively quickly and at low cost compared with the investment required to develop new specialised products. This study focussed on simple key research questions: – What are the barriers to the purchase and use of cell phones by seniors with impairments? – What are the cell phone features and other factors that affect the use of cell phones by seniors with impairments? – Is it feasible to enable seniors with impairments to overcome these barriers by providing informal individual support?

2. Method 2.1. Procedure There were two major components of the study; an expert review of the purchasing process and available products and a field trial of selected phones with a group of seniors with impairments. Expert Review: In the first part of the expert review one of our research team investigated the way that cell


phones are sold and the experience of purchasing a cell phone and the implications for older people in obtaining products that meet their needs. We then reviewed the accessibility of currently available cell phone products for older people with mild to moderate vision, hearing, and dexterity impairments. To accomplish this we used three sources, the web, retail stores, and direct product testing. All of this data was entered into a spreadsheet with narrative comments. One researcher did all of the technical work while another kept notes on the process and debriefed the technical expert on the experience. Website Review: We started with a series of searches to find any specialized phones that were available through the web. We then reviewed every page of every website of every company that offered service to the Berkeley and Oakland, California area. We also reviewed the manufactures’ web pages of all of the phone handsets sold through all of the local service providers. This included a review of phone specifications and manuals and a trial of any instructional material available. Retail Review: During a three month period from May to July 2007 we then visited (at least twice each) at least two branch retail stores of each local Berkeley or Oakland, CA cell phone provider that had retail store. This included long conversations with retail staff on issues surrounding vision, hearing, and dexterity impairment. Selected Product Testing: We then chose the phones we thought would best meet the needs of disabled elderly. We set a cut off of phones available for no more than US $200 with or without a service plan. We purchased a total of 16 different models of phones, including one child’s and one “senior” model. We also reviewed an array of accessories, including after-market screen magnifiers. Because each manufacturer tends to use different software options we purchased at least one phone from a retail store outlet and one phone online from each local company that offered a no-service contract option. We also purchased at least one phone from each manufacturer the local services offered. We used an audio recorder and the “talk out loud” method to record the process of purchasing, setting up, learning, and using each phone. This technique also called verbal protocols is often used in research on learning. In this case we adapted it to a technique of mindful reflection on one’s own learning [31]. This technique has been used for research on technology design [33,34]. One researcher opened and set up all the phones. She kept a running commentary on the process, describing each step into the audio recorder. This included notes on ev-

151

ery detail from the saw needed to open the blister pack to the impatience of the activation operator to a caller with a speech impairment. Note that one of the authors has a speech impairment. We then checked the actual features of each phone against its advertised features and the accuracy, completeness, and the accessibility of the manuals. Where feasible, we brought our phone to the appropriate company’s retail outlet and asked for help adjusting settings to best meet low vision, hearing, or dexterity issues. We requested help with making and receiving calls and with phonebook and voice mail and took notes on the diversity of instructional strategies used by different companies and individual employees. Field Trials: In a team of two people we conducted a simple field trial of cell phone use with a small sample of people. We used an initial screening interview, a longer initial interview, a series of support visits and encouraging phone calls as needed, and an exit interview. Our research was reviewed by the University of California Institutional Review Board for the protection of human subjects. We employed a descriptive flyer and all participants signed consent forms. All names use are pseudonyms. We kept narrative ethnographic field notes. One researcher did most of the in person show and tell on how to use the phone while the other took notes and did most of the follow-up encouragement calls. Recruitment Screening and Initial Interview: For this study we recruited a sample of 12 (including one couple) people over age 65 with vision, hearing, or dexterity impairments who live in the Oakland and Berkeley area of California. We recruited by posting fliers at a large residential facility with both skilled nursing care and lower level care units. We also posted fliers at a senior centre run by the city and one run by a church. We attended luncheon or other activity groups at these three sites and set up a table with flyers and an array of phones for people to look at. Approximately one quarter of our participants called us. The majority of participants first met us in person. In an initial interview at the recruitment event, or at the respondents’ home if they called us, we asked about their current or prior cell phone use and if they experienced any vision, hearing, or dexterity impairment. After this simple screening conversation and if they were judged by the researchers to be sufficiently impaired we scheduled an initial interview either at the facility where we met them or at their home. This meeting assessed the communication and accessibility needs of participants. We asked about the types of difficulties they experienced and we carried out some basic tests of phone functions e.g. dialling a

152


number, looking at screens, and testing loudness with several different handsets. This took one to three meetings as we brought various phones for them to try. We had them try to execute a call to us and to answer calls from us. Phone Distribution, Support, and Encouragement: Using the information we gathered above we then identified the most appropriate phone(s) we could for each person and brought them to the participants in individual visits, usually in their home. We settled on a phone with the participants’ involvement, distributed it, and informally trained participants on how to use them in one to three visits that varied from a half hour to three hours each, as needed. We did not prepare any specific training materials or curriculum. Our intent was to act as if we were a family member, friend, neighbour, or other untrained trainer. We adjusted sound, screen properties and other customisable features by trial and error until we and the participant agreed we had the best mix available. We helped participants enter a few important phone numbers into their phone book and tested incoming, outgoing, and voice mail calls. We demonstrated battery charging. Participants then used the phone during a minimum one-month trial period with effectively unlimited airtime and as much individual support, problem solving, and encouragement as they asked for or needed in personal visits and frequent calls. If they did not answer our weekly test calls we made a home visit. We readjusted phone settings as needed and repeated test calls as necessary in person. We saw or spoke to them at least once a week. We often sat with them and the manual for their phone making adjustments. Exit Interview: At the conclusion of the trial we again scheduled interviews at facility or home. We interviewed participants about their experience during the whole trial. We gave them the option to retain the handset at the end of the trial as our gift, and we helped them choose a service plan, if needed. Data Recording: Anthropologists Devva Kasnitz and Russell Shuttleworth conducted all interviews and visits together. We made some notes during the events and debriefed each other immediately after the event. We kept extensive observational and conversational field notes. We also kept records of all phone calls. 2.2. Participants Our participants were recruited at senior centers and at an assisted living village in the Oakland and Berkeley area. To be eligible to participate in the study par-

ticipants needed to experience a significant degree of impairment in either vision, hearing or manual dexterity and be interested in learning to use a cell phone. We chose people who did not then have a cell phone or who had an old one they could use only with great difficulty. We excluded people with cognitive problems or memory loss significant enough that they were troubled making appointments with us or understanding our consent materials. We assessed these criteria in the initial screening interview. We recruited 12 participants, including a husband and wife couple to whom we gave one phone. We interviewed an additional five participants who were not recruited, as they did not meet the selection criteria. Participants ranged in age from 66 to 90 years. Eight lived in their own homes, three in an assisted living facility, and one lived with her daughter. Table 1 shows shows the details of the participants recruited for the study.

3. Results 3.1. Expert review Access Features: In practice, due to the rapid change in the consumer cell phone market, the review of features of available products proved to be very difficult. This market volatility meant that when accessible phones appeared on the market they disappeared quickly. Overall, while most of the features disabled elders need are available in mainstream phones, they tend to be scattered over different handsets and not available together in one package. It is not easy to find information about these features and they may be hard to find in handset menus. Limited market phones for younger adults with vision impairment are very complex, have no screen, depend on scrolling audio menus, and as a result are unsuitable for older people. Some features on mainstream products may be key to accessibility and confidence – for example, speakerphones, voice dial, and larger, brighter screens, such as are featured on some high-end products. Available settings on some phones can be very helpful if an older person or their helper is aware of them; for example, the backlight of some phones can be set to remain on permanently, which may be essential for users with impaired vision. Some speakerphones are loud enough for people with hearing loss. Some phones have brighter and clearer screens than others, and allow the default font-size to be increased. However, even when phones do offer useful potential accessibility features, it is often


153

Table 1 Details of Study Participants and Phone Disposition Name Marg

Gender F

Age 66

Hearing poor

Vision poor

Dexterity Poor

Home Own

Ping

F

67

ok

ok

Poor

Own

Angel

F

68

ok

problems

ok

Own, Husband

Diane

F

75

poor

ok

ok

Assisted Living

Judy

F

75

poor/aid

ok

ok

Nursing Home

Ling

F

76

poor

ok

poor

Own, husband

Athena

F

76

poor/aid

ok

ok

Own

Eliza

F

77

poor

ok

ok

Own, Daughter

Martha1

F

79

poor

ok

ok

Own, husband

Peter1

M

80

ok

poor

big fingers

Ben

T

85

poor

ok

poor

Lucy

F

90

poor

ok

ok

Note:

1

Own, Wife Own Nursing Home

Phone type/disposition Samsung SCH-a870 Kept LG CU500 Lost Phone, Would Keep Motorola e815 Kept Jitterbug Returned Jitterbug Returned Samsung 850 Kept Samsung SCH-a870 Kept Sony Ericson S35C2F Kept Jitterbug Returned Jitterbug Returned Jitterbug Returned Jitterbug Returned

Husband and wife living in their own home shared one cell phone. Table 2 Review of Phone Features by Impairment Area

Type of Impairment Vision Impairment

Issues – Limited market phones for “culturally blind” unsuitable for older people due to their complexity. Also phones such as OWASYS do not have a screen and hence cannot readily be used by a sighted person. Many older people with vision impairment will have relatives or caregivers who are sighted and would be unable to support the older person in using these phones. – Mass market phones – some have very good clear screens but most do not maximize font size on anything but phone number dialing. Navigation features are still small. – Backlight on setting is useful for seniors with limited vision but not easy to find in the phone menu. – Some lighted keypads are helpful if characters on keys are large enough and keypad light long enough. Jitterbug phone has very short keypad backlight that was not adjustable and tiny keypad letters. – Voice dial is difficult to use for seniors and often has problems with older voices.

Hearing Impairment

– Jitterbug and newer phones, advertised for hearing impaired may be louder but are marred by poor sound quality and difficult to use volume/speaker phone controls. – The volume and sound quality available from speaker phones is variable and there are no specifications available about this. The only way to determine if a given phone will help an older person with hearing loss is by testing it.

Dexterity Impairment

– Jitterbug phone has big bright buttons with some tactile feedback. – Phones with delicate connections for the charger are a problem as these may easily be damaged by someone with limited manual dexterity. – Many phone keypads are simply too small for men with large fingers to use. This becomes a larger problem for older people with any even minor dexterity limitation. – Rocker style buttons are difficult to use.

difficult to determine which ones. Most phone specifications don’t list all this detailed product information, for example a hard measure of the loudness of a speakerphone, and the design of instructional information for phones is inadequate. Sales staff themselves can

not keep up with changes in their hardware and software inventory. The same model phone will ship with software changes without notice. Table 2 summarizes some of these features. The Business Model: As the US consumer mobile

154


telephone industry is currently organized physical access cannot be separated from programmatic access and business practices. The ergonomic aspects of handset design are mediated by firmware and software that changes faster than does hardware and varies by carrier. The same phone has different capabilities from one carrier to another and from one week to another. Manufacturers and carriers maximize the organization of customisable features on a phone to promote airtime and broadband use, not for ease of basic phone use. For example, IM is easier to get to than is voicemail or missed calls in most default settings. Beyond basic hardware design features of good sound, good screen, large, well-lit and labelled buttons in a familiar arrangement, and standard pin style ports for all accessories, software is more important for access than is hardware. Purchasing issues and market volatility and opacity are huge barriers for senior access. Phone handset manufacturers, software developers, airtime providers, and retail packagers who put it all together, are all separately owned companies that license their products to each other with monopolistic effect. So, for example, almost all phones with voice recognition use the same software, with the same problems understanding nonstandard speech, but each licensee tweaks the system differently. In one case we looked for phones that spoke menus out loud for people with vision issues and found that the company had removed that module of the software routine without even any notice to their own sales staff. The US cell phone industry, in common with other wireless telecommunications providers around the world, is set up so it makes financial sense for a prospective purchaser to choose a service provider, then a tariff (or pricing plan), and then select a handset. That is, the purchaser selects a tariff that specifies the level of network access provided and airtime cost and then selects a handset from the range available for the chosen plan. There are also at least three tiers of industry models based on class that further limit access: models based on cash for people without good credit, regular “residential” plans that feature some free handsets and basic contracts, and “smart phones” that come with business services. Paradoxically, although seniors don’t want “fancy” phones or contracts, the features that most disabled seniors need for access purposes usually require a higher end phone with a long contract and a good credit rating. Furthermore, these phones are expensive but come discounted with only a one year warranty but require a 2 year contract. Insurance is expensive and deductibles high. Taken together, disabled seniors experience an intense pressure to sacrifice access for price.

3.2. Field study Findings from this and a prior initial study [16] show a number of common themes emerging. Older people with disability are interested in mobile phones for the increased security they offer as well as the potential for increased social contact. However, our participants have a common overarching concern about the cost of mobile phones and particularly of airtime. Vision, hearing, manual dexterity, and overall speed all emerge as factors that limit older people’s capacity to access mobile phones. Safety was the first reason our participants stated for wanting to use a mobile phone. Minor emergencies, missed appointments, and the proverbial flat tire are the most commonly mentioned reasons along with the lack of pay phones. However, most participants who have had a cell phone prior to this trial received it as a gift from their adult children. A few people definitely also thought a cell phone would or should be fun. All participants who started the trial completed it. One participant, although she made several calls, never received any and did not answer the phone on any of the times that the researcher called her. With the small numbers of this study it is necessary to be cautious about drawing any conclusions regarding the quantitative findings, however it is interesting that no participants given Jitterbug phones kept them after the trial (See Table 1). Overall the findings show some evidence that it is possible to support older people to use cell phones. For the seniors in our study their use of cell phones was shaped by their living arrangements, goals, wishes, and preferences as well as their physical, sensory, and cognitive capabilities. The brief case studies presented here with pseudonyms illustrate the ways that these factors worked together to influence our participants’ use of cell phones. Case Study Judy: Judy is a retired librarian in her mid seventies. A series of small strokes put her in a nursing home four months ago. She gave up on her old candy-bar phone because of cost and because she could no longer hear it but she carried it with her as clock, alarm, and calendar. We gave her a Jitterbug because it was the loudest phone we had even though it didn’t have the extra features she wanted. She accepted happily. As she tried to learn it we realized that her short-term memory was a problem. She would forget to charge the phone or leave it on for incoming calls and was confused by the “missed call” messages. Although she successfully used the phone several times


it was confusing enough that she stopped carrying it. Her children bought her a new standard phone on their family plan and she returned ours but is not really using the new one either. This makes her goal of going out alone on a bus further away. This participant had the goal of being able to travel alone on the bus and using a cell phone represented a way for her to move towards this goal. She had both cognitive and sensory limitations that significantly restricted her capacity to use a cell phone. For this participant the memory and comprehension demands to use and maintain a cell phone may have made it too difficult a task. Case Study Peter and Martha: Peter, a retired history professor in his early eighties, and Martha, his wife were hopeful that they might find a cell phone he could see, at least with his magnifying glass, and she could hear when turned up high, that would also be economical. They were adapting to a new cordless land line and it proved too much to get used to two new phones at the same time. The Jitterbug voice dial could not recognize Peter’s gravelly voice and had to be recharged too often. Martha found the manual both paternalistically “dumbed down” and incomplete and she apologized for just not studying it enough. They found the steps to answer the phone difficult to master. The Jitterbug seemed paradoxically both too complicated to learn to use and too simple in features to be worth the effort. As they returned it to us Martha joked, “If I want a cell phone, I’ll just get a real one.” But they have no serious intention of getting another cell phone. This case study illustrates a relationship in which both persons have impairments associated with aging and want to share a phone to economize. This means that phones need to be able to accommodate the accessibility needs of multiple users. This case study also indicates some issues associated with the Jitterbug phones. These phones had a number of significant usability issues that impacted on how they were perceived by seniors. The battery life was short and in practice the phone needed to be recharged daily for most of the participants. In order to make the phone controls simple, the company restricts access to aspects of its functionality from the handset. For example, in order to program numbers into the phone book the user has to call an operator to do it for them or to request that the needed software module be sent to them. The older users perceived these restrictions as frustrating and limiting their control and we noticed that the process of telling the operator the names of people for the phone

155

book introduced a lot of typographical errors so that users had to remember that the operator programmed “Joey” as “Zoe.” Case Study Athena: Aged 75 Athena, who is from Greece, lives alone. Her hearing is becoming more of a major problem and she can’t hear her old cell phone at all. During a power failure that knocked out her landline her daughter had to scream at her on the old cell phone. It took us over two hours and two visits to her quiet home to find a phone she could hear without annoying static. This could not have been accomplished in most stores. Athena was very happy to keep the phone although she felt the pay-as-you-go plan was too expensive and she has no interest in a contract. We suggested she consider sharing a plan with her daughter. This final case study illustrates the ways that a sensory impairment, hearing loss, resulted in difficulty in both purchasing and accessing the product. Due to her hearing loss this participant would not have been able to identify a phone that she could use in the shop environment or online. No cell phone specifications tell a consumer which phones are louder than others. However, because she lives alone she had a strong motivation to be able to use a cell phone as an emergency contact with her daughter. Other cases taught us other lessons. Lucy was apologetic that at 90 she is old and dumb. However, her literalness in following the very long Jitterbug recorded instructions to access voicemail made her its most successful user, Finally many participants struggled with basic conceptual issues e.g. the need to turn on the phone to make and receive calls and to press send to make a call. However, by the end of the study most participants could make a basic call.

4. Discussion Market forces now drive basic communication access. The industry is opaque and advocacy very difficult. Planned obsolescence rules. There are several diverging and converging mobile technologies. There are some clever divergent disability niche technologies. “Even though what stands out at first glance is the convergence of functions into single devices, a more careful look reveals the divergence.” [4, Page 25] The choices are greatest for tech-savvy, affluent, blind people with hardware from Japan or Czechoslovakia, and software for smart phones and PDAs [34]. There are phones targeted for children, people with

156


cognitive impairment, people living in the global south, and a few targeted to address the needs of the elderly. Most of these divergent products are marginally successful at best. Meanwhile, mobile device technology is converging with remote medical technology. Soon the mobile phone will replace a dedicated “panic button,” will send blood sugar readings to a central computer and will monitor pacemakers. The main findings of our expert review are that the sales model and handset features of consumer cell phones do not readily accommodate the needs of older people with impairments. The technology exists. It is just not appropriately bundled, even by the companies that target the disabled elderly market. Most access problems can be solved with both more user customisable software and more preset profiles. So, for example, users need to control font style, colour, and size more than wallpaper. They need to control length of ring more than tune. They need to be able to hide menu options they don’t want to use and to reorder others. Once set, a user needs to be able to then lock their customized profile. Also, preset profiles for low vision or hearing loss could be programmed just as a "silent" profile is available. The sales model that starts with choosing a service plan and cost makes it difficult for older people with specific accessibility needs to select a product that will be easy for them to use, especially as seniors are often on limited incomes. The findings of the field study suggest that with intensive support seniors with impairments are able to learn to use cell phones although many struggled with basic concepts [35]. Consistent with other findings [3,5], seniors’ main motivation for having a cell phone is safety although in practice they often used cell phones for smaller issues in their day to day life e.g. when late for an appointment. Seniors’ cell phone use is influenced by their goals and living arrangements as well as their physical and cognitive and sensory capabilities. Consistent with other research [36] we found that seniors perceived the so called “senior phones” negatively, we consider their usability poor, and that they restricted seniors’ choice and control. How could cell phone carriers and handset designers better support older users? Cell phone carriers could work with retailers and distributors to make information available to purchasers about the accessibility features of cell phone handsets, such as a hard measure of the volume output of a speaker more transparent. Australia has recently introduced a provision to ensure that information about a specified set of features that

support accessibility of phone handsets will be publicly available [37]. Carriers and retailers could also have handsets with good support for access for people with impairments available on lower price tariffs with a short or no required contract period or early termination fee. The American Association of Retired People, with its enormous lobby, has taken this up as a policy plank [38]. Another associated US equity issue is the lack of fit between the minimum length of a required service contract, usually two years, and the length of handset manufacturers warranty, usually one year. This leaves consumers purchasing costly insurance with a large deductible, a new phone at full cost, or a simpler, cheaper phone that may not have the same needed features for best access. Older people and their relatives and assistants should be able to identify phones that will meet the needs of seniors with impairments by seeking information online and from disability groups. They may need to talk to retailers in advance to ask if the older person can test drive different phones. With assistance and support during the learning period older people can learn to use cell phones. It may be helpful to identify a friend or relative who they can call and who will call them so that they get practise in using the phone and in solving problems in the early stages. Further studies are needed to understand the specific kind of instructions, training, and support that works best in helping older people to master cell phone use. Learning how we can effectively help older people to use everyday technology has been identified as an important issue that may have broad implications in supporting seniors in living independently [39]. Further work is also needed to understand how instructional materials, training, and support can best be delivered. This could be through automated means such as an online or DVD training package or through direct instruction via community seniors’ organisations. Training older people with impairments to use cell phones could represent a cost effective alternative to providing them with specialised monitoring and alert systems. Moreover compared to specialized alarm and monitoring systems that tend to be focussed on the needs of healthcare providers and government it is likely to be less isolating and support freedom of choice, control and access to the wider environment for older people with impairments [39–41]. In conclusion this exploratory study suggests that if the barriers in the business and sales model and design of current mass market cell phones can be overcome, then older people with impairments, with support, could use off the shelf cell phones as a tool to


support independent living. Transparency is needed all along the chain from advertising, to contracts, to billing. Some issues will only be solved by policy and regulatory systems to enforce and enhance standards and transparency. Other issues will require more research to look at how older adults best learn to use this technology. Our preliminary impression is that something more systematic than ad hoc familial, instruction is needed. Contextual issues are as important as a physical access issue. Disabled seniors need features common to younger disabled people that may represent universal design. The loud speakerphone will assist a business meeting or a busy mother with a wet baby, a communication curb cut. Given the growing numbers of older people with impairments, the majority of whom strongly desire to remain living independently, we believe that this potential merits further attention from researchers, policy makers, and industry. What kind of training is needed and how can it best be delivered – in a retail store – online – in the community? The industry that promotes consumer control and choice, and that offers flexibility, portability, transparency, and an accessible business model will support agency for disabled elderly.

[5] [6]

[7]

[8]

[9]

[10]

[11]

Acknowledgments The authors wish to acknowledge the financial support of Sprint Foundation for the conduct of this research project. Also the Ed Roberts Fellowship in Disability Studies, Institute of Urban and Regional Development, University of California, Berkeley, funded by the National Institute of Disability and Rehabilitation Research Grant H133P020009.

[12] [13]

[14] [15]

References [1] [2]

[3]

[4]

International Telecommunications Union http://www.itu.int/ themes/accessibility/, Accessed June 22, 2009. Counting each individual phone card purchased, European data for 2006 suggests that there are more cell phone subscriptions on average than people. Eurostat http://nui.epp.eurostat. ec.europa.eu/nui/show.do?dataset=tin00060&lang=en, Accessed June 24, 2009. W.W.-L. Wong, Mobile Phones, Aged Homes, and Family Relations in Hong Kong Preliminary Observations, pages 179– 193, in New Technologies in Global Societies. edited by PuiLam Law, Leopoldma Fortunati, and Shashua Yang. 2006 World Scientific Publishing. J.R. Vaughn, Over the Horizon: Potential Impact of Emerging Trends in Information and Communication Technology on Disability Policy and Practice National Council on Disability, December 19, 2006, 55 Pages.

[16]

[17]

[18]

157

H.A. Horst and D. Miller, The Cell Phone: An Anthropology of Communication, Oxford: Berg, 2006. C. Law, The technology in your cell phone wasn’t invented for you. Paper Presented at the Communication Policy and Research Forum 25–26 September, 2006, Sydney, Australia. J. Simpson, Inclusive Information and Communication Technologies for People with Disabilities. Disability Studies Quarterly Winter 2009, Volume 29, No.1 . Point of Care Laboratory, In home monitoring: Unobtrusive continuous in-home monitoring. http://www.bme.ogi.edu/PO CL/unobtrusive continuous.php. Accessed 15/10/2007. P.M.A. Baker, N.W. Moon and A. Bakowski, Access to Wireless Technologies for People with Disabilities: Issues, Opportunities and Policy Options Findings of a Policy Delphi December 2007, Wireless RERC1/Center for Advanced Communications Policy(CACP) Georgia Institute of Technology, 24 pages, http://www.wirelessrerc.org/publications/policybriefs/brief files/Wireless policy delphi 07.PDF. Accessed May 24, 2009. The World Summit on the Information Society, through the Geneva Plan of Action, and the Doha Plan of Action urged Member States to address the special requirements of persons with disabilities in their national e-strategies and encouraged the design and production of ICT equipment and services suited to their needs, including adherence to the Universal Design Principle and use of assistive technologies. A further call was made for research and development to facilitate accessibility of ICT for all, including people with disabilities. http://www.itu.int/wtisd/2008/theme.html, accessed June 23. 2009. C. Baker and K.K. Kim-Sung, Understanding Consumer Concerns about the Quality of Wireless Telephone Service / Research Report July 2003 – Findings from a nationwide survey on consumers’ interest in, awareness and understanding of, and satisfaction with wireless telephone service and service providers are reported in this AARP Public Policy Institute Data Digest. (4 pages) (Publication ID: DD89). K. Brown, Newer phones, older users. Wireless Week, 2005, http://www.wirelessweek.com/article.aspx?id=60. National Statistics. Adult mobile phone ownership or use by age 2001 and 2003: Social Trends. (n.d.), //www.statistics.gov. uk/statbase/ssdataset.asp?vlnk=7202, Accessed October 27, 2007. G.C. Vanderheiden, Thirty-something (million): should they be exceptions? Human Factors 32(4) (1990), 383–396. R. Mace, Definition, The Center for Universal Design web site. Retrieved June 23, 2009 from http://www.design.ncsu. edu/cud/about ud/udprincipleshtmlformat.html, 1997. Also Also see Tusler, Vol 25, No 2 (2005) , Disability Studies Quarterly, Spring 2005, Volume 25, No. 2, , How to Make Technology Work: A Study of Best Practices in United States Electronic and Information Technology Companies. R. Pedlow, Providing accessible cell phones for the elderly disabled: What can we learn from the experience of younger people with disabilities? Paper presented at the Society for Disability Studies Conference, June 2006, Washington DC. T. Nguyen, R. Garrett, A.D. Downing, L. Walker and D. Hobbs, Research into Telecommunications Options for People with Physical Disabilities, Asst Technol 19 (2007), 78–93. N.J. Nelson, Bonnie O’Day and Verizon Wireless Settle Section 255 Formal Complaint Proceeding, 08/27/2004. http://www.icdri.org/News/vzboday.htm. Also see, Guernsey,

158

[19]

[20]

[21]

[22] [23]

[24]

[25]

[26]

[27] [28]

[29]

R. Pedlow et al. / Barriers to the adoption of cell phones for older people with impairments in the USA L., February 12, 2004, New York Times, A new cell phone nods to the needs of the disabled. http://www.nytimes.com/ 2004/02/12/technology/circuits/12phon.html. Accessed May 11, 2009. Rehabilitation Engineering Research Center for Wireless Technologies, 2007, http://www.wirelessrerc.org/. Accessed December 2, 2008. Rehabilitation Engineering Research Center on Telecommunications Access, a partnership between the Technology Access Program of Gallaudet University and the Trace Center of the University of Wisconsin–Madison. http://trace.wisc. edu/telrerc/. J. Sajka, The Information Requirements of Persons who are Blind or Visually Impaired. Capital Accessibility, LLC, Technology Access Program, Gallaudet, page 2. http://tap. gallaudet.edu/emergency/nov05conference/Papers/Sajka.asp. Accessed May 2, 2008. M. Putnam, (Editor.) Aging and Disability: Crossing Network Lines. Springer: NY. 2007. W.C. Mann, S. Helal, R.D. Davenport, M.D. Justiss, M.R. Tomita, and B.J. Kemp, Use of cell phones by elders with impairments: overall appraisal, satisfaction and suggestions, Technology and Disability 16 (2004), 49–57. J. Abascal and A. Civit, Mobile Communication for Older People: New Opportunities for Autonomous Life. EC/NSF Workshop on universal accessibility of ubiquitous computing: providing for the elderly. 22–25 May 2001, Alcacer do sal Portugal. S. Friedman, S. Munoz, B. West, S. Rubin and G. Fried, Falls and fear of falling: which comes first? A longitudinal prediction model suggest strategies for primary and secondary prevention. Journal of the American Geriatrics Society 50 (2002), 1329–1335. L. Brooks and H. Kendig, Understanding community nursing for older individuals and carers, in: Nursing Older People: Issues and Innovations, R. Nay and S. Garrett, eds, Melbourne: MacLennon and Petty, 2004. Jitterbug. Keeping it simple with Jitterbug. http://www. jitterbug.com/. Accessed 15/10/07. M. Drummond, Choosing the Right Cell Phone, How do you decide on the phone that’s best for you? Disabled World – Jan 21, 2009 4:56:47 PM, http://www.clarityproducts.com/ products/listing/item3289.asp?gclid=COLs9LbKgJsCFRd75 QodMQlZdA. Accessed May 2, 2009. Alsp see, http:// assistive-technology.suite101.com/article.cfm/choosing the right cell phone#ixzz0I3zIemIb&C, http://assistive-technol ogy.suite101.com/article.cfm/choosing the right cell phone. accessed June 22, 2009. For example, see http://cellphonesforseniorcitizens.blog.co.

[30]

[31]

[32]

[33] [34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

uk/2009/03/04/cell-phone-for-senior-citizens-motorola-e85683383/. T. Sokoler and S. Svensson, Embracing ambiguity in the design of non-stigmatising digital technology for social interaction among senior citizens, Behaviour and Information Technology 26 (2007), 297–307. E. Katalin, “Please, keep talking”: The ‘think-aloud’ method in second language reading research, http://ludens.elte.hu/˜ deal/pages/novelty/htm2/vol73/elekes.htm. Education Encyclopaedia: Research Methods: Verbal Protocols http://www.answers.com/topic/research-methods-verbalprotocols,. Accessed May 2, 2009. Trace Center. (2006). User needs summary, http://trace.wisc. edu/docs/2006/user-needs-summary/. Accessed July 18, 2007 J. Wagner, G.C. Vanderheiden and M.E. Sesto, Improving the Usability of a Mainstream Cell Phone for Individuals with Low Vision, Journal of Visual Impairment & Blindness 100(11) (Nov/2006), 0145482X. H. Tang and S. Kao, Understanding the real need of the elderly people when using mobile phones. Paper presented to the International conference on inclusive design. Royal College of Art, 5–8 April 2005, London, UK. American Association of Retired Person. n.d. Older persons and wireless telephone use. http://www.aarp.org/research/ utilities/phone/fs116 wireless.html. Accessed May 20, 2009. ACIF (2005). C625: 2005 Information on Accessibility Features for Telephone Equipment. http://www.commsalliance. com.au/documents/codes/C625 (accessed 12/3/2008). C. Baker, Cell Phone Service Bills, Long-Term Contracts, and Complaints. Research Report. AARP Public Policy Institute, June 2006, http://www.aarp.org/research/ utilities/phone/fs130 complaints.html. Accessed 6/26/2009. L. Johnson and E. Moxon, In whose service? Technology care and disabled people: the case for a disability politics perspective, Disability and Society 13 (1998), 241–258. D. Kaplan, J. de Witt and M. Steyaert, Telecommunications and persons with disability: laying the foundation – a report of the first year of the blue ribbon panel on National Telecommunications Policy. Oakland, California: World Institute on Disability, 1990. S. Wildstrom, Kinder, gentler cell phones: Carriers cater to youth – but the elderly need simpler handsets. BusinessWeek/online, Technology & You. September 19, 2005. http:// www.businessweek.com/magazine/content/05 38/b3951049. htm. A. Bowling, S. Seeta, S. Morris and S. Ebrahim, Quality of life among older people with poor functioning. The influence of perceived control over life, Age and Ageing 36 (2007), 310–315.

127


Robots: Assistive technologies for play, learning and cognitive development Al Cooka,∗, Pedro Encarnaça˜ ob and Kim Adamsa,c a

University of Alberta, Edmonton, AB, Canada Faculty of Engineering, Catholic University of Portugal, Sintra, Portugal c Glenrose Rehabilitation Hospital, Edmonton, AB, Canada

b

Abstract. Robots have been widely used in rehabilitation. Among the various applications, robots have been developed to assist children with motor disabilities in play and academic activities. Several studies have shown the efficacy of these robotic tools, not only for allowing children to actively participate in the activities, with direct impact on the development of their cognitive, social, and linguistic skills, but also as a means to assess children’s understanding of cognitive concepts, when standard tests cannot be used due to physical or language limitations. In this paper the use of robots for assistive play is reviewed from the perspectives of rehabilitation engineering and robot design, aiming at defining a set of desirable characteristics for such robots. Commercially available robots are then surveyed in comparison to the defined characteristics to evaluate to what extent they can be used as assistive robots for play, learning and cognitive development. Keywords: Assistive robotics, play, cognitive development assessment, augmentative communication

1. Introduction During typical development children learn cognitive, social, motor and linguistic skills through manipulation of objects, often in the context of play. Because of motor limitations manipulation of objects may be difficult, and the quality of play and learning of skills may be compromised [51]. Robots can facilitate discovery and enhance opportunities for play, learning and cognitive development in children who have motor disabilities [7]. The usage of robots in play contexts can also help to track changes in cognitive development by the child, and may contribute to improved cognitive understanding [13]. Success with robot tasks could be an alternative way for children to demonstrate their understanding of cognitive concepts avoiding the limitations of standardized test administration, such as verbal response or physical manipulation of objects. ∗ Address

for correspondence: Al Cook, PhD., Professor, Department of Speech Pathology and Audiology, 3-79 Corbett Hall, University of Alberta, Edmonton, AB, Canada T6G 2G4. Tel.: +1 780 492 8954; Fax: +1 780 492 9333; E-mail: [email protected].

This paper begins with a brief overview of robotic systems, and then describes a series of applications to rehabilitation, focusing particularly on robots to reveal cognitive skills for children with disabilities. The results of these and other studies demonstrate the positive impact of the use of robotic systems by children who have disabilities and justify consideration of the design requirements for a robot specifically for this population. Combined with our previous experience [12,13,59] in evaluating and developing cognitive skills in children with disabilities through the use of robots, this material forms the basis for a desired set of robot characteristics for play and education applications. In the final section of the paper a review of commercially available robots based on that set of characteristics is discussed.

2. Robotic systems A robot is defined as “An automatically controlled, reprogrammable, multipurpose, manipulator programmable in three or more axes, which may be either fixed in place or mobile for use in industrial


128

A. Cook et al. / Robots: Assistive technologies for play, learning and cognitive development

automation applications.” (International standard ISO 8373). Although this definition emphasizes manipulators for industrial applications, robots can assume different shapes and are widely used in other areas including rehabilitation, the amelioration of physical sensory and cognitive limitations in children and adults with disabilities. Robots can be programmed to exhibit different levels of autonomy with respect to the user. In one extreme, the robot can accept high level commands specifying a task to be accomplished (e.g., get milk glass), and be able to perform that task making whatever decisions are necessary without requesting any human intervention (fully autonomous). At the other end of the scale, the user has direct control over the robot movements (teleoperated). Multiple controls are then necessary to operate the various robot degrees of freedom. These controls might be directly or indirectly accessible (e.g., through a scanning method). For example, to position a robotic arm end-effector in Cartesian space, controls for x, y, z coordinates must be available. Between these two extremes of autonomy (autonomous and teleoperated robot), several levels of autonomy, shown in Table 1, can be defined, Sheridan’s scale [63] being the most widely cited. One of the most commonly used levels in assistive robots is level 4 of autonomy, in which the user merely needs to hit or press and hold a switch to replay a pre-stored movement. Robots can be roughly divided into the following fundamental components (e.g., [56]): Chassis and Energy, Propulsion and Actuators, Environmental Interface, Navigation, Guidance and Control, Communications, and Mission Control (see Fig. 1). Robots can be mobile or stationary. The following description is for mobile robots, but with the necessary adaptations to each component description, the diagram in Fig. 1 can also be applied to stationary robots. Chassis and Energy refers to the structural part of the robot and to the power system on board, usually made of rechargeable batteries. The Chassis defines the robot shape and relates to the mechanical robustness of the robot and payload capabilities. It should be adapted to the environment where the robot will be used (e.g., indoor or outdoor). The Propulsion and Actuators component is responsible for robot movement. It encompasses the motors and the actuators that transform motor rotational movement into translational robot movement. Actuators can be wheels, drive tracks or artificial legs. The type and geometry of the actuators influences the way a robot can move. It can be holonomic, meaning that all degrees of freedom are controllable, or it can be non-holonomic,

Communications

Guidance

Chassis

and

and

Control

Energy

Mission Control

Propulsion Navigation

and Actuators

Environment Interface

Fig. 1. Fundamental robot components.

only being able to control some of the degrees of freedom. Automobiles are usually non-holonomic since the user can only accelerate/brake and change the angle of the steering wheel in order to control the vehicle three degrees of freedom (x, y position and orientation). The vehicle cannot be moved sideways, for example. However, omni-directional wheels can provide sideways control and an additional degree of freedom. Usually actuators are equipped with odometers that measure the distance traveled by the vehicle (e.g., wheels usually have encoders that give the number of revolutions of the wheel since it started to move). This information, together with wheel radius is sufficient to estimate the distance traveled). In order to perceive the environment and physically act on it the robot must include an Environmental interface component: sensing and manipulation. Sensors can be active, providing the energy necessary for information acquisition, or passive, measuring only available energy. Examples of active sensors are rangefinding sensors like sonar or laser beams. Bumpers (touch), or force sensors are passive. Video cameras may have both passive (e.g. image sensor) and active (e.g. auto-focus system) elements. Robotic arms or some kind of gripper can be added to a mobile robot to enable physical manipulation of the environment, like picking up an object. Often these manipulation tasks require that the robot be able to detect and recognize


129

Table 1 Sheridan’s levels of robotic autonomy (Sheridan and Verplank, 1978) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Computer offers no assistance; human does it all. Computer offers a complete set of action alternatives. Computer narrows the selection down to a few choices. Computer suggests a single action. Computer executes that action if human approves. Computer allows the human limited time to veto before automatic execution. Computer executes automatically then necessarily informs the human. Computer informs human after automatic execution only if human asks. Computer informs human after automatic execution only if it decides to. Computer decides everything and acts autonomously, ignoring the human.

objects in the environment through an appropriate sensory system. For example, force sensors may help to efficiently grasp an object. Also, depending on the robot sensory systems, safety issues aimed at protecting the robot, the environment, and the user (e.g. avoid falling down stairs or hitting obstacles) may be implemented. Navigation systems provide current linear and angular positions and vehicle velocities. The position can be known relative to an initial position (using odometer readings), in relation to obstacles (using range-finding sensors) or in a global coordinate frame (absolute position obtained by a Global Positioning System (GPS) or by comparing range-finding measures with a map of the environment). Some localization methods resort to landmarks placed in the environment which are detected by the robot sensory system. Techniques used for object detection and recognition (e.g. using RFID1 tags on objects) are also used for customizing the environment for robot use. Modifying the environment and objects so they can “communicate” with the robot is usually referred as ubiquitous computing or embedded intelligence. Maps of the environment can be known a priori or can be interactively built by the robot as it moves around. After knowing where the robot is, it is necessary to know where to go and how to get there. First guidance systems define a target point; then control laws drive the vehicle from the current position to the target point. Modern control strategies integrate guidance and control, guaranteeing stability and performance of the combined system. The Communications component encompasses visual and auditory feedback to the user and also the communication systems, if any, that convey information on the robot state or sensory information to a base station. Visual feedback can be provided by means of a display or simply by meaningful use of indicator lights; audi1 Radio

Frequency Identification.

tory feedback can be either sounds, playback of prerecorded messages or text-to-speech generation. Currently, wireless communication systems are standard. Finally, the Mission Control component accepts high level commands from a program or from a user directly to specify a given task and coordinates all the components so the robot can execute the task. Therefore, it encompasses the human-robot interface and the software necessary for subsystem control. This software must be able to coordinate several concurrent processes to achieve a particular goal, and it is usually designed under Artificial Intelligence or Hybrid Systems2 frameworks. The human-robot interface should match the robot user needs. Most probably the robot will be used by non-technical persons thus an intuitive control language should be developed, preferably a graphical one or possibly natural speech. The human-robot interface may also have additional features depending on the characteristics of the users, e.g. children with disabilities. It is at the mission control level that different levels of autonomy are implemented. Table 2 indicates the degree of sensing, feedback and controllability that the robot must have for each of Sheridan’s ten levels of autonomy. Controllability refers to the complexity of the control unit used to obtain robot movement ranging from one button, to a sub-set of functions, to controls for all the degrees of freedom. There are some grey areas between categories in the table. For example, although one could argue that a robot with level one of autonomy (teleoperated) need not be programmable, in order to implement several levels of autonomy in the same system the robot should be fully programmable. Table 2 is a guide to the characteristics that are necessary or desirable for a robot to have in order to function at a given level of autonomy. Thus, typical assistive robots operating at level 4, must have range, cliff, touch 2 Hybrid systems are those that include continuous and discrete states.

Table 2

Level of Autonomy

Sensing Contact Range Ciff Touch Force Image finding 1 Computer offers no assistance; human does it all. ≈ ≈ 2 Computer offers a complete set of action alternatives. ≈ ≈ 3 Computer narrows the selection down to a few choices. × × × ≈ × 4 Computer suggests a single action. × × × ≈ × 5 Computer executes that action if human approves. × × × ≈ × 6 Computer allows the human limited time to veto before auto× × × ≈ × matic execution. 7 Computer executes automatically then necessarily informs the × × × ≈ × human. 8 Computer informs human after automatic execution only if × × × ≈ × human asks. 9 Computer informs human after automatic execution only if it × × × ≈ × decides to. 10 Computer decides everything and acts autonomously, ignoring × × × ≈ × the human. √ Legend: × Must have; ≈ Convenient; One of the items marked must be available. ≈

≈

≈

≈

≈ ≈ ≈ ≈ ≈ ≈ √ √ √

√ √

√

√

√

≈

≈

≈

≈

Auditory Recorded Speech Sound message Synthesis ≈ ≈ √ √ ≈ √ √ ≈ √ √ ≈ √ √ ≈

Feedback

√

√ √ √ √

Visual Lights Display

Hardware/software necessary to implement each of Sheridan’s levels of autonomy

≈

≈ ≈ ≈

×

×

×

×

× × × ×

Controllability Controls Sub-set for of Single all DOF controls control × √ √

130 A. Cook et al. / Robots: Assistive technologies for play, learning and cognitive development


and image sensing to keep from getting into inoperable situations (e.g., falling off of a table) and they must have either visual display or auditory message (preferably both) as feedback to the user/programmer. 3. Robots to support therapy, activities and participation Robotic systems have been widely applied to rehabilitation. Several applications are for children and are described in the next sections. Upper extremity prostheses and exoskeletons are beyond the scope of this paper. Section 4 describes robotic applications to reveal cognitive skills for children with disabilities. The most important and relevant robotic system components from each rehabilitation application will be discussed in Section 4.5. 3.1. Robots for physical therapy The rationale behind applying robots to physical therapy is that “robot-aids not only are more efficient in delivering certain routine physical and occupational therapy activities, but also provide a rich stream of data that assists in patient diagnosis, customization of the therapy, and maintenance of patient records (at the clinic and at home).” [38]. Rehabilitation robots are typically stationary and have articulating parts that guide the child through required movements with control over resistance, speed, and number of repetitions (e.g. [37]). Exploring the plasticity of the brain, the goal oriented repetitive movements that the robots are able to induce may contribute to the re-learning of the movement by the patient, while saving physical and occupational therapists’ time. The Lokomat, a robotic assisted treadmill, has been used therapeutically to improve gait speed, endurance and standing and walking performance in children with cerebral palsy [8,36,47]. 3.2. Robots as personal assistants The goal of applications with robots as personal assistants is to provide manipulation aids to people with motor impairments and or intellectual disabilities, assisting with several everyday functions such as eating or personal hygiene [14]. Personal assistants can be robotic arms which can be stand alone assistive technologies (see e.g. [68] or integrated in a wheelchair [62]). More recently, mobile autonomous platforms with and without robotic arms have been developed to assist elderly and people with disabilities in their homes (see e.g. [16, 55,50]).

131

3.3. Assistive mobility Robotic systems have been applied in assistive mobility to develop power wheelchairs (see e.g. [14], chapter 12, and the references therein) with environment sensors and control systems that enable these wheelchairs to become more autonomous (e.g. [52]). Research has been carried out to design mobile robots that young children can use [10,73]. 3.4. Robots for social integration With the development of the Artificial Intelligence research field, new kind of robots showing aspects of human-style intelligence has emerged. These socially interactive robots are able, to some extent, to perceive and interact with their environment, and have been used to promote social integration. Several studies have been conducted to establish the usefulness of robots in autism therapy. People with autism have impaired social interaction, social communication, and social imagination [18]. Robots could be helpful when human intervention is a barrier to learning, as might be the case with autistic children. Also, it is hypothesized that the “social” relationship the autistic child might develop with the robot can then be transferred to humans. Both stationary robots which imitate human face expressions and gestures, and mobile robots that interact with children trough movement have been developed. Please refer to [18,49]. For a survey on socially interactive robots see [21]. Recently, research has been done addressing the more general problem of developing robots that can be children’s partners or playmates (e.g. [32,7]). 3.5. Using robots to aid functioning by children with disabilities The pioneering work of Seymour Papert [57] showed that robots can enhance motivation and provide a test bed for “learning by doing”. For children with disabilities, robots can provide a means to engage in play and academic activities that involve the exploration and manipulation of the environment [12]. Robots have been used successfully to allow children to participate in school-based tasks that would otherwise be closed to them. A prototype interactive robotic device was used for play by two groups of children, four in pre-school (2 to 4 years old) and five in elementary school (5 to 9 years old), all having moderate to severe physical impairments, and five also with cognitive delays [33,

132


42,43] adapted the Manus arm for use by children with cerebral palsy (CP) by altering both the physical control of the robot and the cognitive tasks required for control. The robot was used for various pick and place academic activities with six participants, 7 to 29 years old, all of whom had CP. The Handy 1 Robot, originally designed as a feeding aid, was adapted for use in a drawing task to allow children to complete assignments with minimal assistance in class alongside peers [64]. A specially designed robot for access to science lab activities was trialed with seven students aged 9 to 11 years who had physical disabilities [28]. Access to the science and art curricula for students, aged 10 to 18 years, who had arthrogryposis, muscular dystrophy, and CP was evaluated with a multi-purpose workstation called the ArlynArm [19]. Harwin et al. [25] describe a robot workstation for use in Education of the physically limited based on the low-cost commercial SCARA robot. Clinical trials with the SCARA involving stacking and knocking down toy bricks, sorting articles, and playing the Tower of Hanoi game, showed that the children enjoyed using the robot and were able to achieve tasks otherwise impossible for them. In the PlayROB project [40] a dedicated robot system which supports children with severe physical impairments for interaction with Lego bricks was developed. A first set of trials was conducted with three able-bodied children (between 5 and 7 years old) and three disabled children (between 9 and 11 years old; child 1 – multiple disabilities; child 2 – tetra paresis; child 3 – transverse spinal cord syndrome). According to the authors, “In general, most of the children enjoyed playing with the system and the goal to make autonomous play for children with physical disabilities possible has been fully achieved.” (page 2899). Upgraded versions of the robot system were then used in a multi-centre study involving children with and without disabilities to investigate possible and estimated learning effects. The encouraging results of this study are reported in [40]. Tsotos et al. [69] present a research project aiming at building a robotic system to access and manipulate toys. The focus of the research in this project has been on the vision system because that’s the greatest technical challenge in the authors’ opinion. The problems faced by children with mobility impairments were addressed in the initial stages of the project (see e.g. [5]). All the above applications in this section use robots as therapeutic or enabling tools. However, as stressed in Section 1, observing children using robots can also provide a means to assess their cognitive development. Section 4 is dedicated to robots for manipulation which are used in scenarios designed to test children’s demonstration/development of cognitive skills.

4. Using robots to reveal cognitive skills for children with disabilities The potential of using robots to reveal cognitive skills for children with disabilities has been referred to by other authors (e.g. [23,71]). Research projects in our group (see references 1–4, 11–15, 59, 65) have focused on cognitive skills associated with robot use in three ways. First, robot studies have been designed to require specific cognitive skills by the child. Generally, constrained in some way, these investigations have revealed underlying cognitive skills that may have been undetected or not easily measured by more traditional means. Second, environments of discovery have been developed in which children with disabilities are encouraged to explore and problem solve using robots. These unconstrained studies have provided a platform on which children can demonstrate a variety of cognitive skills. Finally, there have been studies of robot use by young typically developing children. Three of the robots used in these studies are shown in Fig. 2. 4.1. Means end causality and tool use by infants Infants typically develop the concept of tool use in which they understand the relationship between objects and use one object to obtain another (e.g., using a stick to extend reach to push a toy) by age 8 months [70]. In order to determine if this concept would also apply to the use of a robot, the MiniMover robot arm (Fig. 2) was used with young children with and without disabilities aged 6–18 months in a direct control task in one dimension [11]. The MiniMover is a half human scale robotic arm with six degrees of freedom (shoulder, elbow, wrist and base rotation, and wrist flexion and extension). It is designed for table top use in an open loop control mode. The robot arm held a cracker. When the child pressed a switch the arm moved it closer, and when the child released the switch the arm stopped moving. Reaching for the cracker and then pressing the switch when the cracker was out of range was taken to mean the child was using the robot as a tool to bring the cracker closer. This conclusion was also supported by observed behaviors such as point of visual regard (e.g., looking at the arm, then looking at the switch, then pressing the switch, then looking back at the arm expecting it to move) and affect (smiling, crying, laughing to indicate level of enjoyment or distress) during task [11]. The use of behavioral analysis such as this has also been reported by Dautenhaun and Werry [17] who called them “micro-behaviors”. Three


133

Fig. 2. Robots used in studies with children who have disabilities.

typically developing and six developmentally delayed children participated in the study in a pre-school setting. An infant development scale (Bayley Developmental Scale) was used to asses the cognitive age of the children [6]. All typically developing children whose age was greater than eight months used the robot arm as a tool and younger children did not. All children with disabilities who had a cognitive age greater than eight months also used the arm as tool. 4.2. Sequencing in a constrained task In order to evaluate the ability of young children to carry out a three step sequencing task, the Rhino robot, shown in Fig. 2, was used by twelve children 5– 10 years old who had severe physical disabilities [12]. The Rhino robot is an industrial robotic arm with the same degrees of freedom of the MiniMover. It is also designed for table top use but has built in closed loop control systems for each joint and can also be controlled in Cartesian space coordinates. Children controlled the robot using single-switches. None of the participants were able to engage independently or with another child or adult without some adaptation in co-operative play in which objects of various materials (e.g., sand, water, macaroni)” are placed in containers of various types and then dumped out for the sensory (auditory, tactile and visual) feedback that occurs. All of the participants had experience using single switches to operate toys and to access computer games, but for many of the

children, consistent switch access was generally not established, and it was difficult to assess cognitive and language skills using standardized measures. A large tub of dry macaroni noodles was used as the medium for burying objects. There were three tasks for the child. The first task involved pressing switch 1 to cause the robot to dump the macaroni from a glass. The second task had two switches each controlling one step: (1) press switch 2 to dig an object out of the macaroni, and (2) switch 1 to dump the macaroni and object. The third task consisted of a three step procedure for the child: (1) press a 3rd switch to position the robot arm for digging (using up to 8 increments of movement requiring multiple presses of the switch), (2) press the switch 2 to dig an object out of the macaroni, and (3) press the switch 1 to dump the macaroni and object. Goal Attainment Scaling (GAS) [35] was used to evaluate the participants’ level of achievement in these three tasks. Five levels were assigned: expected result (value = 0), two better performance levels (+1, +2) and two worse (−1, −2). Examples of GAS Scales are shown in Table 3. All twelve of the participants were able to independently control at least two switches in the sequence. Seven of the children independently used all three switches and one used three switches with some prompting. Teachers initially thought that the researchers had overestimated the skills of the children in selecting them for this project. At the end of the study, teachers were surprised at the level of accomplishment

134

A. Cook et al. / Robots: Assistive technologies for play, learning and cognitive development Table 3 Examples of goals used in goal attainment scaling Goal attainment scaling guide Functional Carry over Unexpected gains in Understands “your classroom activities turn” and responds appropriately

Goal +2 Best Expected Outcome

Operation Controls 3 switches with auditory and visual prompts

+1 Better Than Expected Outcome

Controls 2 switches with a 2 step scoop with out assistance

Understands turn taking with auditory and visual prompts

Shows more interest in classroom activities. Eg. Increase in vocalization, words, attention span, and enthusiasm

0 Expected Outcome

Controls 2 switches intentionally with out assistance

Anticipates and responds accordingly to her turn

Increased interaction, and becomes excited when teacher speaks about the robot session

−1 Less Than Expected Outcome −2 Worst Expected Outcome

Controls 2 switches with prompting Controls one switch intentionally

Anticipates the outcome of the task Enjoys interaction with the instructor during the trial

No change

of the children. A set of open-ended questions were also used with the teachers to provide insight into child’s social and academic performance before and after using the robot. The primary themes from the teachers were: Reactions: – “[student] Smiled and got excited when robot mentioned in class or at home.” – “Robot gave [student] something to look forward to and become excited about.” – Children’s reactions to robot were very positive – Robot tasks were more motivational (generated more interest and excitement) than single switch tasks with toys, appliances and computer-based activities Communication: – “had more vocalizations in class, and was more interactive after robot use.” – “[Student] used new symbols in class and interaction increased.” Confidence – “[student’s] confidence and interaction increased, he looked forward to the sessions.” – “On the way to robot [student] anticipated what was going to happen; her ability to control robot increased [student’s] self esteem.” This study is reported in more depth in [12].

Shows less attention, is more passive in the classroom, and anticipates session when she sees the therapy staff

4.3. Discovery and problem solving in unconstrained tasks In the previous two studies [12,13] the tasks were constrained due to the nature of the robots and the type of control required. The robots were also expensive ($1500 to $10,000US), making replication in schools and children’s home difficult. Lego MindStorms3 robots cost approximately $300US and provide a very flexible platform for evaluating how children use robots. Namely, one can easily build robots with different shapes, stationary or mobile. Small motors may be used for propulsion or to drive moving parts. Several sensors (e.g. touch, light, sound) are available to provide information about the environment. These robots are equipped with a microprocessor that allows for programming different levels of autonomy. Additionally, they can be remotely controlled via infrared signals. For our studies, the commercial remote controller was adapted to enable the control of the robots using singleswitches. The Lego Invention4 “roverbot” vehicle and robotic arm (Fig. 2) were used to determine if low cost robots could provide a means by which children with severe disabilities can demonstrate understanding of cognitive concepts [13]. Both constrained and uncon3 http://mindstorms.lego.com/en-us/default.aspx. 4 http://mindstorms.lego.com/.


135

Table 4 Robot skills related to development of cognitive skills. (From [13] Skill

Definition for robot use

Age considerations* (typically developing children) NA

0

No interaction

Child displays no interest in the robot or its actions

1

Causality

2

Lego Robot Examples NA

Understanding the relationship between a switch and a resulting effect

< 3 action is in switch, tried to use disconnected switches > 4 yrs understood switch made robot move

Use switch to drive robot, knocking over blocks with robot, drawing circles on paper by holding a switch down and turning robot

Negation

An action can be negated by its opposite

4 yrs: begin to understand that switch release stops robot

Releasing switch to stop robot

3

Binary Logic

Two opposite effects such as on and not on

5–6 yrs: understood 2 switches with opposite effects.

2 switches turning robot right/left, or go and stop

4

Coordination of multiple variable Spatial conceptsmultiple dimension

Movement in more than one dimension to meet a functional goal

age 5: Could fine tune a movement by reversing to compensate for overshoot, etc

Moving roverbot to a specific location in two dimensions

5

Symbolic Play

Make believe with real, miniature or imaginary props [51]

6 yrs: Child ID action in robot not switch, planning of tasks is possible

Interactive play with pretense, i.e. serving at tea party, exchanging toys with friends , pretending to feed animals all using robot

6

Problem solving

Problem solving with a plan – not trial and error, Generation of multiple possible solutions

7 yrs. Designed robot and thought about coordinated effects, planning was possible, Can understand simple programs and debug

Changing strategies to solve a problem such as avoid an obstacle, Changing task to meet the child’s own goal, simple programming

From Forman (1986).

strained tasks were utilized in a study with different levels of autonomy that allowed free play and discovery by ten participants who ranged in age from four to ten. Their disabilities were primarily cerebral palsy and related motor conditions with widely variable motor, cognitive and language abilities. All had complex communication needs and were non-speaking. Initially, participants used single switch activation to activate pre-stored movements such as a robot dancing, knocking over a stack of blocks or drawing circles on a large piece of paper. This established that the child had an understanding of causality and the function of the switch in controlling the robot. For participants who demonstrated understanding of robot control, four switches were used in an unconstrained discovery task in which the child controlled the roverbot to turn (left/right) and move (go/stop). For some children the switches were accessed with hand movement and for others it was a combination of hand and head movement. In order to characterize and evaluate the cognitive skills being demonstrated by the participants during the unconstrained use of the robot, a comparison to robot use by typically developing children was used. In a study of three to seven year olds using a

RobotixTM robot5 construction kit, children demonstrated five problem solving skills: causality, spatial relations, binary logic, the coordination of multiple variables, and reflectivity [22]. The specific robot skills achieved in each of these areas varied with the age of the children. Stanger and Cook [65] studied typically developing children, one to three years of age, using a Hero 2000 robot6 in a series of increasingly cognitively complex tasks. Cognitive skills investigated included causality and the use of sequencing two and three switches to carry out a task. Two and three year old children consistently demonstrated causality, while the youngest children (16 months) were inconsistent in this task. Only the three year old children were able to complete the two step sequencing task. None of the children completed the three step sequence successfully. Based on these studies of typical children’s use of robots, the set of robot tasks shown in Table 4 was developed. Each task requires cognitive skills of varying levels of complexity. A child’s performance on these tasks, which are progressively more cognitively chal5 Robotix

discontinued, but see http://en.wikipedia.org/wiki/ Robotix (toys). 6 Hero 2000 discontinued, but see http://www.hero2000robots. com/9501.html.

136


lenging, provides a proxy measure of cognitive understanding by children with disabilities performing robot tasks by comparison to typically developing children at different ages. The results of the study involving ten children with disabilities is summarized in Table 4 and reported in more depth in [13]. In a recent study, eighteen typically developing children aged three, four and five years used a Lego robot to complete tasks based on the cognitive concepts of causality, negation, binary logic and sequencing [59]. All of the participants understood causality. The four and five year old children grasped the concept of negation, but the three year olds had more difficulty understanding this concept. Most of the 4 and 5 year old participants succeeded at the binary logic (left and right) task. Forman found that only children older than four were able to understand binary logic. This may have been due to Forman’s use of one rocker switch with two directions of movement whereas this study used two separate switches for each direction located spatially on the left and right side of the forward switch. None of the three year olds were able to consistently use a two step sequence to accomplish a task. Four year olds displayed greater understanding of the sequencing task than younger children, while five year olds had no problem in accomplishing the task. This study verified that the cognitive skills listed in Table 4 develop at the ages shown for typical children. 4.4. Integrating communication and robotic manipulation Children who have motor limitations are sometimes also limited in communication. These children may use Speech Generating Devices (SGDs) to meet some of their communication needs. SGDs are stand alone or computer based electronic devices that produce digitized or speech output in response to selections made by the child using a variety of input methods including typing, head pointing or scanning [14]. One of the challenges faced by these children is the degree to which use of the SGD isolates them from other activities including play and academics. For example, an SGD is generally placed directly in front of the child and the child has to turn away from it in order to play. Since much of play and selected portions of the academic curriculum involve manipulation of real objects, integrated systems have been developed so children can communicate and control Lego robots using the same device and access method. Many SGDs have the capability to learn infrared commands, for instance to control televi-

sions and DVD machines. Since the Lego Mindstorm robots are infrared controlled, they can be controlled from SGDs. New generation of Lego Mindstorms and SGD have Bluetooth capability, but that version has not been tested to date. Using communication devices to control the Lego robots addressed several limitations observed in previous Lego robotic play studies. For example, it can be difficult to find the six switch access sites required for control of a three degree of freedom robotic arm for children with severe motor disabilities, and several participants could only use single or dual switches. Thus, using their communication device, these children are able to control the robot via scanning. In addition to scanning, the use of an augmentative and alternative communication (AAC) device opens up the possibility of other alternative access strategies such as manipulation of a cursor through head or eye pointing. Hence, the main difference in the robotic system from the previous Lego robot studies was in the Human-Robot Interface. Two pilot studies of an integrated system have been undertaken [1,15]. In the first study, an integrated communication and robotic play testing platform underwent usability testing and iterative design with professional experts [15] and children with and without disabilities (in preparation). The experts and older children (5 years old) were able to teleoperate the roverbot, but the younger children (3 years old) were only able to control the robot by pressing one switch to initiate a program of a sequence of movements. The results showed that children prefer to do activities using the robot rather than directing another person to do it and that they will spontaneously talk using the communication device during play. The testing platform provided a means to examine the best ways to present information (pages, links, symbols) for finger-pointing users, but requires testing with scanning users. In the second study, a commercially available communication device was used by a single participant (a 12 year old girl who has Cerebral Palsy and uses two switch scanning) to examine the feasibility of controlling Lego robots for academic activities [1,2,4]. This study was useful to establish that it is feasible to control Lego robots via an AAC device for social studies and math activities. With systems such as these, children can demonstrate and develop manipulative, communicative and cognitive skills in an integrated way.


4.5. Robot system components for rehabilitation applications The robot system components shown in Fig. 1 have unique properties in each of the rehabilitation applications discussed in this and the previous sections. For example, in the case of robots for physical therapy, as personal assistants or to provide assistive mobility, the Chassis and Energy component should be specifically designed for its intended use. For other rehabilitation applications, since payloads are generally small and the robots are generally used in indoor controlled environments, the Chassis and Energy and Propulsion and actuators components do not provide major design challenges. Communications generally takes the form of feedback to the user regarding the task and the status of the robot. Most pediatric applications in rehabilitation have limited communication with the child, but they may have significant communication with a user interface remotely accessing the robot for the researcher/therapist/caretaker. This component is also often employed to download programmed tasks into the robot. The level of autonomy (Table 1) has a direct effect on the Environmental Interface, Navigation, and Guidance and Control robot system components. The greater the autonomy, the more dependence there is on the design and implementation of these elements. For most of the applications described here, these components play a small role. The exceptions are robots used as personal assistants, modern assistive mobility devices, or robots for social integration where these components become more important. The Mission Control component is essential for defining and executing particular tasks in a variety of rehabilitation settings. Since it encompasses the human-robot interface, particular attention should be dedicated to this component so the robot can be easily used by a broad range of persons, with and without disabilities (e.g. therapists, caretakers). If the intended users are persons with special needs, appropriate accessible control interfaces and selection methods should be considered (see [14], chapter 7, for a discussion on human-robot interface for persons with disabilities).

5. Characterization of assistive robots for playing and cognitive assessment Commercially available robots were applied in the research using robots to reveal cognitive skills described in the previous section. Most of the compo-

137

nents in Fig. 1 were already implemented and appropriate for the applications. The only customization necessary was modification of the Mission Control System in order to provide accessible control interfaces for children with special needs and programs to carry out the specified tasks. However, the use of commercially available robots poses limitations on the various play and education scenarios where robots could be useful for children with disabilities. For example, the large and robust Rhino robot7 was expensive and required specialized programming. The small but inexpensive Lego robots8 were fragile and required frequent minor adjustments. Small robots also have limited payloads and cannot be implemented for functional tasks with actual play or academic objects. Limited environmental sensing and navigation capabilities may limit the degree of autonomy that can be achieved. Often the Mission Control component is limited to simple commands or short programmed tasks and not suited for more complex scenarios such as automatic adjustment of the degree of autonomy according to the child’s performance. Based on a literature review conducted by the authors in 2008 and on the experience gathered from previous robot studies, the desired characteristics of a robot specifically designed for assistive manipulation in play and school by children who have disabilities are discussed in this section. Literature on design of socially interactive robots and manipulation aids for adults was also surveyed and the concepts that were consistent with our play and education goals were also included. Kemp et al. [34] give an overview of the present challenges in developing robots that perform useful work in everyday settings. 5.1. Design considerations for a children’s robot The main considerations in designing a robot for assistive play and education are: intended use, technological characteristics, aesthetics, and economic considerations (Fig. 3). 5.1.1. Intended use As described by the Human Activity Assistive Technology (HAAT) model [14], the design of any assistive technology system should start by defining the Activities in which the user needs/wants to engage and the Contexts in which those activities will take place. 7 http://www.rhinorobotics.com/. 8 http://mindstorms.lego.com/.

138


Technological

Intended use

characteristics

Robot

Aesthetics

Economic considerations

Fig. 3. Robot design forces.

In the present case, the activity is robot-assisted manipulation that allows children with motor disabilities to engage in play and academic activities providing a tool for exploring, discovering and altering the environment. Ideally the robot will be flexible enough to allow for a wide range of activities. Activities should be developed only by considering user needs and preferences, not by constraints of any specific technological solution [60]. The third aspect of the HAAT model approach is a consideration of the skills the person is capable of for participating directly in the activity and for controlling the interface to the robot. The envisioned activities, contexts and anticipated human skills should then determine the required technological capabilities and characteristics of the robot. 5.1.2. Technological characteristics A desired robot characteristic is robustness [29,41, 49]. A robot that is robust will also be more reliable. When working with children who have disabilities, the need for reliable and accurate robots is essential since failures frustrate and disengage users [12,26,30, 39,41]. Also, robot inaccuracies in performing a given task must be compensated by human intervention, thus limiting independent use of the system by the child. Robustness issues should be taken into consideration when designing each of the robot components shown in Fig. 1. Safety is a key issue when designing robotic systems to be used by children [29,30,39,44,48,49,53]. The robot cannot, in any situation, place the child in danger. Moreover, “the acceptance of robots for health-care applications has been slowed by safety concerns” [44, p. 298]. Lees and LePage [44] also discuss the tradeoff between cost and safety. Michaud et al. [48] specifically refer to the problem of avoiding small parts. Safety has direct implications in the design of the robot chas-

sis since it should incorporate passive elements that prevent injury to the child in case he or she comes in contact with the robot. Also the sensory system should enable the robot to detect and anticipate dangerous situations. Finally, the mission control system should implement safety procedures that override any other vehicle operation. It is mostly at the mission control level that special attention is necessary when developing robots for children with disabilities, since this component encompasses the human-robot interface and the coordination of all robot subsystems to accomplish a specific task. Software design for robot control has its base in the Human Robot Interface (HRI) research field and is dependent on considerations from different areas of expertise such as psychology, physical and occupational therapy, speech and language pathology, artificial intelligence, and computer science. Goodrich and Schultz [24] provide a comprehensive survey of human-robot interaction. Human-robot interfaces should be intuitive and accessible for non-technical individuals [26]. The development of “standard control software to enable the use of the same programs across robotic systems” is also important [30, p. 150]. A rehabilitation robot for children should be usable by children who have a variety of disabilities, should have easily learned operation, and should include simple and comfortable access to input devices [29,44]. The interface software should provide easy, transparent access to the capabilities of the robot empowering the user and giving a sense of effective control over the system and environment [29]. Additionally, the robot should interact in a natural manner with the user and robotic systems, operation should be easily learned by non-technical users and children [49] and provide appropriate feedback. Kronreif and Prazak-Aram [41] aimed at “plug & play” operation for their robotic system. Teaching the robot through the child’s own body motion is a natural way for children to program the robot [58]. However, teaching by the child’s movement is more directed to socially interactive robots or physical therapy than to assistive robots for children who have motor disabilities to augment physical movement of objects. In principle, the HRI should provide full control over the assistive technology (e.g., allowing all degrees of freedom of a mobile robot to be controlled) so the user can be fully in charge of the activity. In practice, due to motor or cognitive impairments, it might be necessary for a portion of the control to be taken over by the robotic system (e.g. Levels 2–10 of Table 1). This may limit


the freedom of the child (e.g., to move the robot from one point to another the user might only be able to select pre-specified destinations using a scanning method and the system would plan the trajectory and drive the robot through that trajectory, avoiding obstacles if they occur). Progressing through levels of autonomy (e.g., from Level 10 to Level 4 or 5 of Table 1) allows for the gradual introduction of the control of the assistive technology. Decisions to move from one level to another can be based on skills such as those shown in Table 4 since each higher level in Table 4 requires greater autonomy by the child and less by the robotic system. The robotic system should then have a high degree of autonomy in the first trials, but can gradually release control to the user as he or she masters the system. If possible, after the learning process, the user should be able to fully control the robotic system (Level 1). Thus, software that allows for the implementation of different levels of autonomy that match the child’s performance level must be developed. Ideally the robot would automatically adapt its level of autonomy according to user performance. Several authors report the use of different levels of autonomy in their robotic systems (cf. [10, 13,25,27,29,53]). Some of these robotic systems offer more than one level of autonomy, but none of them automatically adapt. Future development of robotic systems that automatically adapt the level of autonomy according to user performance would increase the ability of the robot to help the child develop autonomy as skills increased. For a discussion on the user interface for an assistive robot based on considerations from robotics, cognitive science, human factors engineering, and humancomputer interaction fields, please refer to [66]. Most important concepts to a successful human-robot interface design, like visibility (of the controls), conceptual mapping (between the controls and the actions), feedback, modeless interaction, and error recovery or reversible actions, are common to the design of everyday things [54,66]. Stanger’s paper also addresses the evaluation of specific interface designs, stressing the importance of involving the user in the design loop at early stages. Other desirable technological characteristics for a child’s robot include: portability, so the robotic system can be used in a place the child knows, thus reducing distractions and anxiety [26,53,72]; availability of a logging system to record every play session in order to assess possible learning effects from the robot use [40]; and usage of vision sensors, essential to enable self adaption of the robot to changes in the environment [25, 44].

139

5.1.3. Aesthetics Robotic systems for children must be appealing to the child and significant others [28]. Aesthetics in the design of assistive technologies for children has been studied for augmentative communication devices and robots separately. Light et al. [45] examined popular toys for young children to identify potential designs that might improve the appeal of AAC systems. Recognizing that current AAC devices are designed by adults, Light et al. [46] conducted a study in which six children without disabilities were asked to design low-tech AAC prototype systems to obtain an indication of the children’s preferences. The children’s inventions differed significantly from the designs of current AAC technologies, namely they incorporated multiple functions (e.g.,communication, social interaction, companionship, play, artistic expression, telecommunications), provided dynamic contexts to support social interactions with others, and made use of bright colors, lights, transformable shapes, popular themes, humor, and amazing accomplishments to capture interest, enhance appeal, build self-esteem, and establish a positive social image [46]. The AAC systems designed were seen as children’s companions and were easily personalized to reflect the user’s age, personality, attitude, interests, and preferences. Bumby and Dautenhahn [9] conducted a study with thirty eight children between the ages of seven and eleven, divided in groups of nine to ten, to identify how they perceived robots and what type of behavior they may exhibit when interacting with robots. It was found that children see robots as having geometric forms with human features in their faces and feet for walking, placed them in familiar settings and social contexts, and attributed free will to them. Despite the familiarity with the technology all groups showed a tendency to overestimate the capabilities of the robots. The robots should be appealing to attract children’s attention [12, 49,61,67]. Examples of ways to make robots appealing to children are the use of bright colors, replication of well known children’s themes (e.g., cartoon or book characters), incorporating amusing movements or actions, and allowing for easy personalization to match the child’s preferences. Balance must be obtained between creating an attractive robot and distracting the child from the tasks by too many cosmetic features. These considerations have impact on the design of the chassis and communications components of the robot.

140


5.1.4. Economic considerations Along with safety, cost is one of the most frequent limitations of rehabilitation robots cited in the literature [20,26,30,31,39,44]. There is a cost-performance tradeoff, industrial-grade robots are robost but expensive and cheaper robots designed for education are inexpensive but not robust. Two observations found in the literature are: “poor cost-to-performance ratios have been the major weakness of the ‘functional aid’ applications for robotic/mechatronic systems” [20, p. 24], and “in education, professionals have concentrated on finding ways of forcing cheap robots to barely meet their needs rather than developing robotic systems that are truly well suited for educational purposes” [44, p. 298]. In fact, cost is still the limiting factor in developing robots for children who have disabilities that can be widely used in school or home scenarios. 5.2. Characteristics of commercially available robots A web based survey of commercially available robots was carried out by the authors in 2009 in order to compare available technology with the desired robot characteristics. Ready-to-operate robots, kit robotic systems, and mobile platforms designed for educational robotics development were surveyed. Search criteria, reflecting those characteristics we considered fundamental to the robot were: a) Mobile robots with manipulation capabilities or with the option of adding manipulation capabilities, to allow for a wide range of play activities; b) Dimensions compatible with table play activities so it is possible to use the robot in different play contexts (school or home) while seated in a wheelchair or supportive seating system; c) Wireless and omnidirectional user robot interface to avoid cables and “shadow” control zones where the robot looses communication with the user interface; d) Cost less than 600 Canadian dollars. This constraint is included in order to make the robot widely available. A graphic programming language is desirable so non-technical persons, including children, who have disabilities can program the robot. Table 5 presents the best candidates among the robots considered. Prices are based on the web search and thus are approximate. In order to operate every robot listed it is necessary to program it first. All robots have a potential payload of at least 200 gr. All systems re-

quire the design of a customized human-robot interface to make them accessible to children with special needs. If a particular robot accepts commands via infrared signals, off the shelf assistive technology devices can be used to control the robot such as switches, communication devices or a computer. New assistive technology devices incorporate Bluetooth technology thus also allowing for the control of robots that accept commands via Bluetooth. Table 5 characterizes each robot and relates it to the robot components shown in Fig. 1. Additional detail has been added to some of the component categories shown in Fig. 1. Due to the cost constraint, only Mindstorms NXT from Lego or Robot Explorer from Fischertechnik (first two robot columns of Table 5) meet the design specifications without any additional technical development at the robot level. The previous Lego studies reported on earlier in this paper were conducted with the earlier infrared controlled Lego Mindstorms robot, not the currently available Bluetooth version shown in Table 5. A wide range of sensors comes with these systems and the manufacturers provide a visual programming environment usable by non-technical persons, including children. Various robot configurations can be built with the kits, allowing for some customization. Also, being a mass product designed for children, they are appealing, easy to operate, and documentation and parts are readily available. Assuming the child will not have the opportunity to disassemble the robot and come into contact with small robot pieces, these robots do not raise any safety issues. Drawbacks of these systems include robustness and reliability. Robots built out of these kits are fragile and usually do not perform consistently due to construction weaknesses (e.g., gears can easily get misaligned). Most of these “technical” problems can easily be solved by nontechnical helpers, e.g. peers or caretakers. However, these problems can be both confusing and frustrating to the child and can reduce their independence. Robot columns three to eight of Table 5 list educational robotic systems or mobile platforms that require some mechanical/electronic development in order to meet the robot design specifications. In all, no intervention is needed in the Chassis and Energy components and the open hardware architecture allows for hardware customization. All but the Protobot Kit from VEX robotics require the addition of manipulation capabilities. A set of sensors is provided with every listed robot, but additional sensors may be needed. Sensors can be easily added by a technical developer resorting to off-the-shelf kits (see e.g. www.phidgets.com). An


141

Table 5 Commercially available robots Robot Mindstorms NXT LEGO

VEX Robotics

Non-technical

Informed user

requirements

Non-technical

Programming

Non-technical

Operational

$312,50

requirements

CAD

Protobot Kit

Fischertechnik

$350,00

Non-technical

Non-technical

$500,00

iRobot Create

$287,50

Expert

Informed user

Interactive C

A4WD1 Robot

Arobot-P1

Lynxmotion

Arrick Robotics

$525,00

$500,00

Expert

Expert

Informed user

Informed user

IntelliBrain-Bot

Robot Kit V2.0

Ridge Soft

Inex Robotics $370,50

$550,00

Expert

Expert

Informed user

Informed user

control

Bluetooth remote

Environment Interface

Robot remote

Mission control - Human-robot inertface

Cost

Charact.

Robot Explorer

control using

IR Remote

appropriate

control

software

Cliff and touch

Odometers,

sensors, wheel

negative Light, sound,

temperature

range-finding

coefficient resistor,

(sonar) and touch

photoresistor,

sensors

sonar, color

Line sensor,

drops, distance Limit switch,

traveled, angle

bumper switch,

displacement,

robotic arm

omnidirectional

range-finding 3 obstacle

Odometers

detection sensors

Switch-IR reflector, wheel encoders

sonar), 2 infrared photoreflector

IR sensor (for

sensor, track

sensors

commands or

sensor

sensors (IR and

Virtual Walls) Lights (LEDs,

Communications

function

Lights (LEDs,

programmable), Loudspeaker, display

function

speaker

3 ligths, buzzer

programmable),

(possible to

Piezo speaker

16x2 LCD

sound output

record up to 16

transducer

sequences of

n

Structure

Chassis

configuratio

Operation

notes)

Assembling

Assembling

Assembling

Assembling

Assembling

Assembling

Assembling

necessary

necessary

necessary

necessary

necessary

necessary

necessary

Variable

Variable

Fixed

Fixed

Fixed

Fixed

Variable

Fixed

Size

43.2x31.4x20cm 20x15x20cm

Not available

(arm all the way down pointing

33cm diameter

27.9x31.8x12cm

25.4x25.4x12.5cm

base plate:

Not available

16x6cm

Weight

forward)

~350gr

Not available

Not available

2.9kg

1.8kg

1kg

Not available

Not available

IR Proximity detector ($37.50),

RF Data Link

Accessories

($37,50), ROBO I/O extension ($150,00), Control Set (joystick

color selection,

IR detector

Upper deck

Home base

($18.75), range-

($72.50),

sensor, LED

Transmitter &

($87.50), remote

finding (sonar)

Expansion kit (light

output, infrared

Receiver kit to

control ($31.25),

sensor ($45.00),

(microprocessor)

Pro Software

metal detection,

sensor ($25.00),

Interface

($250,00), ROBO

Sound detection,

tracker line

($200,00), ROBO

allow human

virtual walls

operator control

($50.00), light

($162.50)

sensor ($10.00 for two)

remote control),

touch sensor ($12.50), accelerometer ($31.00), force sensing resistor

Sound + Lights package, Motor sets

sensor, pyroelectric

piezo speaker, tilt

detector, switch

infrared motion

input, temperature

sensor,

sensor, magnetic

temperature sensor and mechanical parts)

($68.75)

field sensor, light detector, CMUCAM2 vision

($6.25), gripper

system

kit ($90.00),

(prices not

remote controller

available)

link

Manufacturer

Picture

($25.00)

mindstorms.lego.com

www.fischertechnik.com

www.vexrobotics.com

store.irobot.com

www.lynxmotion.com

www.arrickrobotics.com

www.inexglobal.com

www.ridgesoft.com

142


interesting accessory is available for iRobot Create: it is an infrared beam emitter that creates a virtual wall that the robot will not cross. That can be used to limit the robot workspace, providing an extra safety feature. Excluding the case of iRobot Create, these robots are not widely available and thus the probability of product discontinuation and after-sale support should be carefully assessed. Although the necessity of customization always raises final product robustness and reliability issues, the educational robotic systems have the potential of performing more consistently over time. If well designed, namely the human-robot interface, they can be as easy to operate and program as the Lego or Fischertechnik systems. Table 5 shows that there are no commercially available robots that are 100% suitable for use by children with disabilities. All described robotic systems require a specific HRI for these children. This can either be a commercial product (e.g. Big Jack from Gewa)9 or adaptation of the standard human-robot interface [12]. IROMEC (http://www.iromec.org, accessed in February 3, 2009) is an example of a project aiming at development of a robotic system specifically designed for children with disabilities. In summary, robotic systems based on Lego or Fischertechnik building blocks can be successfully used as assistive technology devices for assistive play. Being widely available at a reasonable price and being easy to operate by non-technical users, it is conceivable that parents and schools may provide them to their children for long term use. However, it is necessary to adapt the hand held remote controller so people with disabilities can use it. The human-robot interface for the Lego or Fischertechnik robots can be built from commercially available devices thus not preventing children with special needs from using them. Taking advantage of the international communities around Lego or Fischertechnik, multiple research groups can write robot programs and building instructions and make them available over the internet.

6. Conclusions There are several general results that have been noted in all of the studies related to cognitive function and development described above. One of the most important is that overall teachers’ and parents’ percep9 http://www.gewa.se.

tion of the competence of the children increased after successful use of the robots. Universally, the children enjoyed using the robots and anticipated the robot sessions. The use of robots also gave the children a chance to demonstrate a range of cognitive skills while also providing a versatile tool for presentation of tasks, problems and learning opportunities to the child. The insight into cognitive skills provided by the use of the robotic system provides a means of avoiding the limitations of standardized test administration, such as verbal response or physical manipulation of objects. Integration of communication and robot control in play and education activities enhances participation and interest for the child and is effective in providing a means for children to demonstrate integrated manipulative, communicative and cognitive skills. The success of these studies stresses the importance of children to have access to Assistive Robots for assistive play and education. The characterization of rehabilitation robotic systems provides a framework for the consideration of the suitability of commercially available robots for use by children who have disabilities. The Mindstorms NXT from Lego or the Robot Explorer from Fischertechnik may meet the needs of rehabilitation applications, but further development in the area of assistive robots is needed to address the limitations of these commercial devices.

Acknowledgments The Rhino and Lego robot projects were funded by the Stollery Children’s Hospital Foundation, Edmonton Alberta, Canada. Funding for the integrated AAC and play project was provided by the Glenrose Research Foundation, the University of Alberta Endowment Fund for the Future, and the Alberta Economic Development Medical Device Development Program, Edmonton, Alberta, Canada. Elaine Holtham at Aroga in Vancouver, B.C., loaned a Vanguard for research purposes during the Robot control via AAC study. The following colleagues participated in various aspects of these studies: Max Meng, Robin Adkins, Johanna Darrah, Joanne Volden, Norma Harbottle, Cheryl Harbottle, as well as approximately 20 graduate students in speech-language pathology, psychology and biomedical engineering. The work of Pedro Encarnaça˜ o was done during a sabbatical at the University of Alberta and at the Glenrose Rehabilitation Hospital, and was supported in part by a FCT Fellowship.


References [1]

[2]

[3]

[4]

[5]

[6] [7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

K. Adams, J. Yantha and A.M. Cook, Lego Robot Control via a Speech Generating Communication Device for Play and Educational Activities, RESNA Annual Conference, 2008, Washington, DC. K. Adams, J. Yantha and A.M. Cook, Lego Robot Control via a Speech Generating Communication Device for Operational and Communicative Goals, International Society for Augmentative and Alternative Communication – 13th Biennial ISAAC Conference, 2008, Montreal, QC. K. Adams and A.M. Cook, Using an Augmentative and Alternative Communication Device to Program and Control Lego Robots, RESNA Annual Conference, 2009, New Orleans, LA. K. Adams and A.M. Cook, Use of an integrated augmentative communication and robotic play prototype device by children with and without disabilities, 2009 (in preparation). A. Andreopoulos and J.K. Tsotsos, A Framework for Door Localization and Door Opening Using a Robotic Wheelchair for People Living with Mobility Impairments, Robotics Science and Systems (RSS) 2007 Manipulation Workshop: Sensing and Adapting to the Real World, June 30, 2007, Atlanta. N. Bayley, Manual for the Bayley Scales of Infant Development, New York: Psychological Corporation, 1969. S. Besio (ed.), Analysis of Critical Factors Involved in Using Interactive Robots for Education and Therapy of Children With Disabilities, Italy: Editrice UNI Service, 2008. I. Borggraefe, Robotics-Assisted Treadmill Therapy Improves Walking and Standing Performance in Children and Adolescents with Cerebral Palsy, International Neurorehabilitation Symposium, February 2009, Zurich, Switzerland. K.E. Bumby and K. Dautenhahn, Investigating Children’s Attitudes Towards Robots: a Case Study, Proceedings of CT’99: Third Cognitive Technology Conference, August 1999, San Francisco. R. Ceres, J.L. Pons, L. Calderón, A.R. Jiménez, and L. Azevedo, A Robotic Vehicle for Disabled Children, IEEE Engineering in Medicine and Biology Magazine, Nov/Dec 2005. A.M. Cook, P. Hoseit, K.M. Liu, R.Y. Lee and C.M. ZentenoSanchez, Using a Robotic Arm System to Facilitate Learning in Very Young Disabled Children, IEEE Transactions on Biomedical Engineering 35(2) (1988), 132–137. A.M. Cook, B. Bentz, N. Harbottle, C. Lynch and B. Miller, School-Based Use of a Robotic Arm System by Children with Disabilities, IEEE Transactions on Neural Systems and Rehabilitation Engineering 13(4) (2005), 452–460. A.M. Cook, K. Adams, J. Volden and N. Harbottle, Using Lego Robots to Estimate Cognitive Ability in Children who have Severe Physical Disabilities, Disability and Rehabilitation: Assistive Technology, 2009, in press. A.M. Cook and J.M Polgar, Cook & Hussey’s Assistive Technologies: Principles and Practice, 3rd ed., Philadelphia, PA: Elsevier Inc., 2008. M. Corrigan, K. Adams and A.M. Cook, Development of an Interface for Integration of Communication and Robotic Play, RESNA Annual Conference, 2007, Pheonix, AZ. P. Dario, E. Guglielmelli, C. Laschi and G. Teti, MOVAID: a Personal Robot in Everyday Life of Disabled and Elderly People, Technology and Disability 10 (1999), 77–93. K. Dautenhahn and I. Werry, A Quantitative Technique for Analyzing Robot-human Interactions, Proceedings of the 2002 IEEE International Conference of Intelligent Robots and Systems, October 2002, Lausanne, Switzerland, pp. 1132–1138.

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

143

K. Dautenhahn and I. Werry, Towards Interactive Robots in Autism Therapy. Background, Motivation and Challenges, Pragmatics & Cognition 12(1) (2004), 1–35. S.P. Eberhardt, J. Osborne and T. Rahman, Classroom Evaluation of the Arlyn Arm Robotic Workstation, Assistive Technology 12(2) (2000), 132–143. R. Erlandson, Applications of Robotic/Mechatronic Systems in Special Education, Rehabilitation Therapy, and Vocational Training: A Paradigm Shift, IEEE Transactions on Rehabilitation Engineering 3(1) (1995), 22–34. T. Fong, I. Nourbakhsh and K. Dautenhahn, A Survey of Socially Interactive Robots, Robotics and Autonomous Systems 42 (2003), 143–166. G. Forman, Observations of Young Children Solving Problems with Computers and Robots, Journal of Research in Childhood Education 1(2) (1986), 60–73. E.P. Goldenberg, Special Technology for Special Children, Computers to Serve Communication and Autonomy in the Education of the Handicapped Children, Baltimore: University Park Press, 1979. M. Goodrich and A. Schultz, Human–Robot Interaction: A Survey, Foundations and Trends in Human–Computer Interaction 1(3) (2007), 203–275. W.S. Harwin, A. Ginige and R.D. Jackson, A Robot Workstation for Use in Education of the Physically Handicapped, IEEE Transactions on Biomedical Engineering 35(2) (1988), 127–131. W. Harwin, R. Gosine, Z. Kazi, D. Lees and J. Dallaway, A Comparison of Rehabilitation Languages and Software, Robotica 15 (1997), 133–151. R. Howell, S.K. Damarin and P.E. Post, The Use of Robotic Manipulators as Cognitive and Physical Prosthetic Aids, RESNA Annual Conference, 1987, San Rose, California. R. Howell and K. Hay, Software-Based Access and Control of Robotic Manipulators for Severely Physically Disabled Students, Journal of Artificial Intelligence in Education 1(1) (1989), 53–72. R. Howell, K. Hay and L. Rakocy, Hardware and Software Considerations in the Design of a Prototype Educational Robotic Manipulator, RESNA Annual Conference, 1989, New Orleans, Louisiana. R. Howell, S. Martz and C. Stanger, Classroom Applications of Educational Robots for Inclusive Teams of Students with and without Disabilities, Technology and Disability 5 (1996), 139–150. R.D. Howell, Robotic Devices as Assistive and Educational Tools for Persons with Disabilities, in: Handbook of Special Education Technology Research and Practice, D. Edyburn, K. Higgins and R. Boone eds, Knowledge by Design, Inc., 2005. A.M. Howard, H.W. Park and C.C. Kemp, Extracting Play Primitives for a Robot Playmate by Sequencing Low-Level Motion Behaviors, Proceedings of the 17th IEEE International Symposium on Robot and Human Interactive Communication, August 1-3, 2008, Technische Universität München, Munich, Germany. G. Karlan, S. Nof, N. Widmer, I. McEwen and B. Nail, eds, Preliminary Clinical Evaluation of a Prototype Interactive Robotic Device (IRD-1), ICAART 88, 1988, Montreal, Quebec. C.C. Kemp, A. Edsinger and E. Torres-Jara, Challenges for Robot Manipulation in Human Environments. Developing Robots that Perform Useful Work in Everyday Settings, IEEE Robotics & Automation Magazine, March 2007. T.J. Kiresuk, A. Smith and J.E. Cardillo, eds, Goal Attainment

144

[36]

[37]

[38] [39]

[40]

[41]

[42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

A. Cook et al. / Robots: Assistive technologies for play, learning and cognitive development Scaling: Applications, Theory and Measurement, Hillsdale, NJ: Erlbaum, 1994. A. Koenig, M. Wellner, S. Köneke, A. Meyer-Heim, L. Lünenburger and R. Riener, Virtual Gait Training for Children with Cerebral Palsy Using the Lokomat Gait Orthosis, Studies in Health Technology and Informatics 132 (2008), 204–209. H.I. Krebs, B.T. Volpe, D. Williams, J. Celestino, S.K. Charles, D. Lynch and N. Hogan, Robot-Aided Neororehabilitation: A Robot for Wrist Rehabilitation, IEEE Transactions on Neural Systems and Rehabilitation Engineering 15(3) (2007), 327– 335. H.I. Krebs, http://meche.mit.edu/people/research/index.html? id=130, accessed January 12, 2009. G. Kronreif, B. Prazak, S. Mina, M. Kornfeld, M. Meindl and M. Fürst, PlayROB – Robot-Assisted Playing for Children with Severe Physical Disabilities, Proceedings of the 9th IEEE International Conference on Rehabilitation Robotics, June 28–July 1, 2005, Chicago, IL. G. Kronreif, M. Kornfeld, B. Prazak, S. Mina and M. Fürst, Robot Assistance in Playful Environment – User Trials and Results, Proceedings of IEEE International Conference on Robotics and Automation, April 10–14, 2007, Rome, Italy. G. Kronreif and B. Prazak-Aram, Robot and Play – from Assistance to Mediation, ACM/IEEE Human-Robot Interaction Conference (HRI08), March 12–15, 2008, Amsterdam, the Netherlands. H. Kwee, J. Quaedackers, E. van de Bool, L. Theeuwen and L. Speth (eds.), POCUS Project: Adapting the Control of the MANUS Manipulator for Persons with Cerebral Palsy, International Conference on Rehabilitation Robotics (ICORR), July 1–2, 1999, Palo Alto, CA. H. Kwee, J. Quaedackers, E. van de Bool, L. Theeuwen and L. Speth, Adapting the Control of the MANUS Manipulator for Persons with Cerebral Palsy: an Exploratory Study, Technology and Disability 14(1) (2002), 31–42. D. Lees and P. LePage, Will Robots Ever Replace Attendants? Exploring the Current Capabilities and Future Potential of Robots in Education and Rehabilitation, International Journal of Rehabilitation Research 17 (1994), 285–304. J. Light, K. Drager and J. Nemser, Enhancing the Appeal of AAC Technologies for Young Children: Lessons from the Toy Manufacturers, Augmentative and Alternative Communication 20(3) (2004), 137–149. J. Light, R. Page, J. Curran and L. Pitkin, Children’s Ideas for the Design of AAC Assistive Technologies for Young Children with Complex Communication Needs, Augmentative and Alternative Communication 23(4) (2007), 274–287. A. Meyer-Heim, I. Borggraefe, C. Ammann-Reiffer, S. Berweck, F.H. Sennhauser, G. Colombo, B. Knecht and F. Heinen, Feasibility of Robotic Assisted Locomotor Training in Children with Central Gait Impairment, Developmental Medicine and Child Neurology 49 (2007), 900–906. F. Michaud, A. Duquette and I. Nadeau, Characteristics of Mobile Robotic Toys for Children with Pervasive Developmental Disorders, Proceedings of the IEEE Conference on Systems, Man, and Cybernetics, October 5–8, 2003, Washington, DC, pp. 2938–2943. F. Michaud, T. Salter, A. Duquette and J.F. Laplante, Perspectives on Mobile Robots Used as Tools for Pediatric Rehabilitation, Assistive Technologies, Special Issue on Intelligent Systems in Pediatric Rehabilitation 19(1) (2007), 21–36. F. Michaud, P. Boissy, D. Labonté, H. Corriveau, A. Grant, M. Lauria, R. Cloutier, M.A. Roux, D. Iannuzzi and M.P. Royer, A Telementoring Robot for Home Care, Technology and Aging,

[51] [52]

[53]

[54] [55]

[56]

[57] [58]

[59]

[60]

[61]

[62]

[63]

[64]

[65]

[66]

[67]

selected papers from the 2007 International Conference on Technology and Aging, 21, 2008. C.R. Musselwhite, Adaptive Play for Special Needs Children, San Diego, CA: College-Hill Press, 1986. P. Nisbet, J. Craig, P. Odor and S. Aitken, ‘Smart’ Wheelchairs for Mobility Training, Technology and Disability 5 (1996), 49–62. S. Nof, G. Karlan and N. Widmer, Development of a Prototype Interactive Robotic Device for Use by Multiply Handicapped Children, ICAART 88, 1988, Montreal, Quebec. D. Norman, The Design of Everyday Things, Basic Books, 2002. H. Nguyen, C. Anderson, A. Trevor, A. Jain, Z. Xu and C.C. Kemp, El-E: an Assistive Robot that Fetches Objects from Flat Surfaces, The Robotic Helpers Workshop at HRI’08, March 12, 2008, Amsterdam, The Netherlands. P. Oliveira, A. Pascoal, V. Silva and C. Silvestre, Mission Control of the MARIUS Autonomous Underwater Vehicle: System Design and Implementation, and Sea Trials, International Journal of Systems Science, special issue on Underwater Robotics 29(10) (1998). S. Papert, Mindstorms: Children, Computers, and Powerful Ideas, New York: Basic Books, 1980. C. Plaisant, A. Druin, C. Lathan, K. Dakhane, K. Edwards, J.M. Vice and J. Montemayor, A Storytelling Robot for Pediatric Rehabilitation, Proceedings of ASSETS’00: ACM Conference on Assistive Technologies, November 13–15, 2000, Arlington, Virginia. L. Poletz, P. Encarnaça˜ o, K. Adams and A.M. Cook, Robot Skills of Preschool Children, RESNA Annual Conference, 2009, New Orleans, LA. B. Robins, N. Otero, E. Ferrari and K. Dautenhahn, Eliciting Requirements for a Robotic Toy for Children with Autism – Results from User Panels, 16th IEEE International Conference on Robot & Human Interactive Communication, August 26– 29, 2007, Jeju, Korea. B. Robins, E. Ferrari and K. Dautenhahn, Developing Scenarios for Robot Assisted Play, Proceedings of the 17th IEEE International Symposium on Robot and Human Interactive Communication, August 1–3, 2008, Technische Universität München, Munich, Germany. J.C. Rosier, J.A. van Woerden, L.W. van der Kolk, B.J.F. Driessen, H.H. Kwee, J.J. Duimel, J.J. Smits, A.A. Tuinhof de Moed, G. Honderd and P.M. Bruyn, Rehabilitation Robotics: the MANUS Concept, Proceedings of the Fifth International Conference on Advanced Robotics, June 19–22, 1991, Pisa, Italy, vol. 1, pp. 893–898. T. Sheridan and W. Verplank, Human and Computer Control for Undersea Teleoperators, MIT Man-Machine Systems Laboratory, 1978. J. Smith and M. Topping, The Introduction of a Robotic Aid to Drawing Into a School for Physically Handicapped Children: a Case Study, British Journal of Occupational Therapy 59(12) (1996), 565–569. C.A. Stanger and A.M. Cook, Using Robotics to Assist in Determining Cognitive Age of Very Young Children, Proceedings of the 12th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, November 1–4, 1990, Philadelphia, PA, pp. 1911–1912. C.A. Stanger, Designing the User Interface for an ElectroMechanical Assistive Manipulator, ASEL Technical Report #ROB9403, May 1994. M. Topping and K. Smith, The Development of Handy I. A robotic System to Assist the Severely Disabled, Technology


[68]

[69]

[70]

[71]

and Disability 10 (1999), 95–105. M. Topping, An Overview of the Development of Handy 1, a Rehabilitation Robot to Assist the Severely Disabled, Journal of Intelligent and Robotic Systems 34 (2002), 253–263. J.K. Tsotsos, G. Verghese, S. Dickinson, M. Jenkin, A. Jepson, E. Milios, F. Nuflo, S. Stevenson, M. Black, D. Metaxas, S. Culhane, Y. Ye and R. Mann, PLAYBOT: A Visually-guided Robot to Assist Physically Disabled Children in Play, Image & Vision Computing Journal, Special Issue on Vision for the Disabled 16 (1998), 275–292. I.C. Uzgiris and J. McV. Hunt, Assessment in Infancy: Ordinal Scales of Psychological Development, Urbana: University of Illinois Press, 1989. S. Weir, S. Russel and J. Valente, Logo: An Approach to Educating Disabled Children. Action-oriented Learning Has

[72]

[73]

145

Dramatic Benefits, BYTE Publications Inc 9 (1982), 346–360. I. Werry, K. Dautenhahn, B. Ogden and W. Harwin, Can Social Interaction Skills Be Taught by a Social Agent? The Role of a Robotic Mediator in Autism Therapy, Proceedings of CT2001: Fourth International Conference on Cognitive Technology: Instruments Of Mind, August 6–9, 2001, University of Warwick, United Kingdom; Springer Verlag, Lecture Notes in Computer Science, subseries Lecture Notes in Artificial Intelligence. C. Wright-Ott, The Transitional Powered Mobility Aid: A New Concept and Tool for Early Mobility, in: Pediatric Powered Mobility, Developmental Perspectives, Technological Issues, Clinical Approaches, J. Furumasu (ed.), Arlington, VA: RESNA, 1997.

117


Robot skills and cognitive performance of preschool children Linda Poletza , Pedro Encarnaça˜ oc,1, Kim Adamsa,b and Al Cooka,∗ a

Faculty of Rehabilitation Medicine, University of Alberta, Edmonton, AB, Canada Glenrose Rehabilitation Hospital, Edmonton, AB, Canada c Faculty of Engineering, Catholic University of Portugal, Sintra, Portugal b

Abstract. Several studies have demonstrated the potential of robots as assistive tools for play activities. Through the use of robots, children with motor impairments may be able to manipulate objects and engage in play activities as their typically developing peers, thus having the same opportunities to learn cognitive, social, motor and linguistic skills. Robot use can also provide a proxy measure of disabled children’s cognitive abilities by comparing their performance with that of typically developing children. This paper reports a study with eighteen typically developing children aged three, four and five years to assess at which ages the cognitive concepts of causality, negation, binary logic, and sequencing are demonstrated during Lego robot use. Keywords: Assistive robotics, play, cognitive development assessment

1. Background During typical development, play activities provide an opportunity for children to learn cognitive, social, motor and linguistic skills through the manipulation of objects. Children who have movement disorders may have difficulty manipulating objects, thereby compromising the quality of play and learning of skills [15]. It can be difficult to ascertain the developmental level of children with motor disorders since many standardized tests are difficult to use and interpret with this population due to the requirement to use speech or fine motor control, or both (children with motor disorders frequently also have speech disorders). Consequently these children may be perceived as being more developmentally delayed than they actually are. Robots provide an opportunity for them to choose how to inter1 The work of Pedro Encarnac ¸ a˜ o was done during a sabbatical at the University of Alberta and at the Glenrose Rehabilitation Hospital, and was supported in part by a FCT Fellowship. ∗ Address for corresponence: Al Cook, PhD., Professor, Department of Speech Pathology and Audiology, 3-79 Corbett Hall, University of Alberta, Edmonton, AB, T6G 2G4, Canada. Tel.: +1 780 492 8954; Fax: +1 780 492 9333; E-mail: [email protected].

act with their environment, to exert some control over the activity, and to manipulate three-dimensional objects. Play-based manipulation using robot tasks can also provide a method for children to demonstrate their understanding of cognitive concepts. Robots have been used successfully in a number of studies to allow children with disabilities to participate in play and engage in school-based activities. Preschool and elementary school children with moderate to severe physical impairments, and cognitive delays participated in manipulative tasks using a robot [11]. Children with cerebral palsy (CP) used an adapted Manus arm for various pick and place academic activities [13, 14]. The Handy 1 Robot, originally designed as a feeding aid, was adapted for use in a drawing task to allow children to complete assignments with minimal assistance in class alongside peers [19]. A specially designed robot for access to science lab activities was trialed with seven students aged 9 to 11 years who had physical disabilities [10]. Access to the science and art curricula for students, aged 10 to 18 years, who had arthrogryposis, muscular dystrophy, and CP was evaluated with a multi-purpose workstation called the ArlynArm [7]. Robot use allowed control over component actions of complex sequences to complete aca-


118

L. Poletz et al. / Robot skills and cognitive performance of preschool children

demic science tasks [16]. Children with disabilities used a robot workstation based on the low-cost commercial SCARA robot for stacking and knocking down toy bricks, sorting articles, and playing the Tower of Hanoi game [9]. In the PlayROB project [12], a dedicated robot system which supports children with severe physical impairments in their interaction with standard toys was developed. A first set of trials was conducted with three able-bodied children (between 5 and 7 yrs old) and three disabled children (between 9 and 11 yrs old). The majority of children were able to use the robot independently and appeared to enjoy the activity. Upgraded versions of the system were then used in a multi-centre longitudinal study involving children with and without disabilities. Results showed that children were able to progressively master the robot, playing autonomously with high concentration and enjoyment, even for long periods of time. Additionally, improvement on child’s spatial perception was reported [12]. There is an ongoing Playbot project, aimed at building a robotic system for assistive play using vision as the primary sensor [1,21]. Another project, IROMEC, is investigating how robotic toys can become social mediators and provide opportunities for learning and enjoyment and focuses on the importance of play in child development and the role that robotics can play in enabling play by children who have disabilities [2]. The IROMEC project team has developed a set of play scenarios that serve to set the context for users to be involved in the design process of appropriate robotics activities and hardware. They have identified four types of play: sensory motor play, symbolic play, constructive play and games with rules [18]. A flexible modular mobile robot has been developed by the IROMEC project to accommodate multiple users and play scenarios [17]. The robot can be adapted to play scenarios with three populations of children with disabilities (autism spectrum disorder, intellectual disabilities and severe motor impairment) in three clusters of activities (imitation, actions and coordination, and symbolic play). Most of the previous robot studies carried out with children who have disabilities have focused on compensating for the physical limitations of the child through augmented manipulation. Manipulating an object via a robot is a different task than directly manipulating the object with one’s hand. It is important to understand the cognitive demands that are placed on children who are using robots for functional manipulation. Previous studies have reported the use of robots to demonstrate previously unmeasured cognitive skills,

even in very young children. Disabled and typicallydeveloping children greater than 8 months in age demonstrated the cognitive skill of tool use by using a robot to bring an object closer to them [3]. A multistep structured play task to uncover a hidden toy was carried out by children aged 6-14 who had severe cerebral palsy [4]. The children performed a sequence of tasks by activating one or more switches. Even though the majority of the participants could not be evaluated through standard cognitive measures, teachers noticed differences in overall responsiveness, amount of vocalization and interest (i.e., increased attention to tasks) for children who used the robotic arm,. Overall, these studies demonstrate that using the robots children can reveal skills that had not been previously measured. In order to gain a sense of the cognitive performance level of children with disabilities using robots, performance of typically developing children at varying developmental ages can be used as an informal measure. However, there have not been many studies showing children’s skills in robot use at different ages. Children aged three to seven using a RobotixTM robot construction kit demonstrated five cognitive skills: cause and effect relations, spatial relations, binary logic, the coordination of multiple variables, and reflectivity [8]. The specific skills demonstrated by the children in each of these areas varied with age, i.e., older children demonstrated greater understanding of each concept than did younger children. Stanger and Cook [20] studied typically developing children one to three years of age using a Hero 2000 robot in a series of increasingly cognitively complex tasks. Two questions were asked in a five step protocol. First, does the child use the robot to do something interesting for him (cause and effect)? Second, can the child use a sequence of robot control commands to carry out a task? As in Forman’s study, older children demonstrated greater understanding of each concept than did younger children While Forman [8] and Stanger and Cook [20] are the only studies of which we are aware that specifically looked at typically developing young children’s understanding of robotic skills, the developmental sequence of skills reported in those studies is similar to those described by standard measures of typical cognitive development [22], and in classification schemes such as the World Health Organization, International Classification of Functioning for Children and Youth (ICFCY) [23]. The ICF-CY includes developmental considerations for children in a number of areas. The categories of Mental Functions (included in Body Functions) and Learning and Applying Knowledge (includ-


ed in Activities and Participation) are particularly relevant to the current study. Classifications that are related to the cognitive functions and use of robots include the mental functions of orientation to objects, motivation, attention, organization of psychomotor functions (including goal directed sequences), and basic cognitive functions (e.g. “acquisition of knowledge about objects, events and experiences; and the organization and application of that knowledge in tasks requiring mental activity” [23, classification b163]). Activity and participation classifications in the ICF-CY that relate to work with children and robots include learning through simple actions with single and/or multiple objects, acquiring basic concepts, making decisions among choices and “carrying out simple or complex and coordinated actions as components of multiple, integrated and complex tasks in sequence or simultaneously” [23, classification d220]. With respect to studies showing the robot skills of children with disabilities, we are aware of only one. In a study with children with disabilities, ten children were observed during unstructured robotic play activities to determine if they demonstrated certain cognitive skills. An observation checklist was used that was based on the cognitive skills observed by Forman [8]: Causality, Negation, Binary Logic, Spatial concepts in multiple dimension (i.e., making sequential movements in multiple dimensions), Symbolic Play, and Problem solving) [6]. Note that negation was studied by Forman under cause and effect relations. It was found that even the children who were not testable with standardized tests were able to demonstrate skills with the robot up to the level of sequencing. The children with the most severe cognitive disabilities understood causality but not negation or binary relations. The sequence of skill understanding with increasing age (causality, then negation, then binary relations) appeared to apply to these children as well. However, in this case the progression in skills was related to their cognitive or developmental level, and not necessarily chronological age. In order to use the demonstration of robot skills as a proxy measure of cognitive level, it is necessary to examine more closely at what ages the robot skills emerge in typically developing children. The purpose of the current study was to confirm the ages at which four cognitive concepts (causality, negation, binary logic, and sequencing) are demonstrated during robot use by typically developing children aged three, four, and five years using a Lego robot controlled with multiple switches. The choice of these cognitive tasks was based on two considerations. First,

119

three of the tasks – causality, negation and binary logicwere shown by Forman to be developmentally related, i.e. older children demonstrated greater understanding of each concept than did younger children. He also showed that these three skills formed a developmental sequence with causality preceding negation and negation preceding binary relations in terms of the ages at which children understood each task, both through demonstrated performance and in answers to subsequent questions regarding that performance. The other skills identified by Forman – the coordination of multiple variables, and reflectivity – were characteristic of older children. This is inline with ICF-CY that includes these cognitive skills in “High-level cognitive functions” [23, classification b164]. Secondly, since our focus was on children for whom cognitive assessment was difficult using standardized measures, we focused on the three to five year old age range, which corresponds to the ages at which Forman saw typically developing children demonstrating the lower-level skills. Due to the importance of sequencing in our previous work with children who have disabilities [4,6] and young children without disabilities [20], we included a sequencing task as well. In both the study by Forman [8] and that by Stanger and Cook [20], the developmental progression by age was based on relatively unstructured play activities and observation of the children. We undertook the current study to provide a more controlled and objective look at these skills.

2. Methodology Eighteen typically developing children were included in the study with ages three, four and five years ± 3 months (Table 1). Informed consent was obtained from the parents for each child in accordance with approved ethics guidelines. Parents were asked to complete the Ages and Stages Questionnaire1 to ensure that the child was functioning within the appropriate developmental level. The children used a truck-like Lego roverbot (Fig. 1) to carry out three tasks which tested the aforementioned cognitive skills. Task 1 (causality) required the child to press and hold a switch until the roverbot knocked over a stack of blocks (Fig. 2). In Task 2 (negation) the child was asked to help build the stack of blocks. 1 http://www.agesandstages.com/index.html.

120


Fig. 1. Lego Roverbot robot.

Fig. 2. Task 1 – Causality: Press and hold a switch until the roverbot knocked over a stack of blocks. Table 1 Participant information Age range 3 years (35–38 mo.) 4 years (46–52 mo.) 5 years (62–63 mo.)

Male 2 5 2

Female 3 3 3

They used the same switch as for Task 1, but they were required to stop the roverbot (i.e., release the switch) beside a pile of blocks to allow the investigator to load them onto the roverbot. Then they were required to stop at the original stacked blocks location where the

investigator unloaded the blocks (Fig. 3). The third task involved two stacks of blocks located to the left and right of the original stack with the roverbot placed between them facing away from the child (Fig. 4). The participant was asked to choose a pile (by pointing at it) and then use the roverbot to knock it down. To accomplish that, the child had to use the appropriate one of two additional switches to turn the roverbot 90 degrees left or right (Task 3A - binary logic), and then use the original forward switch to drive the roverbot to knock over the blocks (Task 3B – sequencing of two


121

Fig. 3. Task 2 – Negation: Move and stop (by releasing the switch) the roverbot beside a pile of blocks to allow the investigator to load them onto the roverbot, and then move and stop the robot at the original stacked blocks location where the investigator unloaded the blocks.

Fig. 4. Task 3A – Binary Logic and Task 3B – Sequencing: Use the appropriate one of two additional switches to turn the roverbot 90 degrees left or right (Task 3A - binary logic), and then use the original forward switch to drive the roverbot to knock over the blocks (Task 3B - sequencing of two actions).

actions). At the end of the session, the children were asked to explain what the switches did in order to assess their understanding of the task. The children used the roverbot at their day care setting or at their home, for two 20 minute sessions spaced approximately seven days apart. All of the tasks were performed at both sessions. The number of trials attempted by each child was dependent on how quickly they understood. Each session was videotaped for analysis. The parents were asked to fill out a technology survey questionnaire to assess the child’s previous familiarity with on/off switches and multi-button remote controls. Frequency of use (1 – Never, 2 – Weekly, or 3 – Daily) and how children mastered those controls (1 – Low skill (trial and error), 2 – Medium skill, or 3 – High skill (mastered)) were assessed.

3. Results The results for the three tasks are summarized in Table 2. Table 3 shows the results of the Welch’s t test (p < 0.05) statistical analysis performed to test the relationship between performance of each task and age level. In all statistical tests it was assumed that the data available for each age group constituted random independent samples of a normally distributed population. Variances of each age group population were assumed to be different. All of the children successfully carried out the first task on all trials. In the second task, only one of the youngest participants did not stop on any trial. The others stopped the robot on at least some of the trials. After having the task explained in more detail their per-

122

L. Poletz et al. / Robot skills and cognitive performance of preschool children Table 2 Summary table of the study results

Participant # 8 12 9 16 7 6 14 10 17 3 15 5 13 20 11 4 18 19 Age (months) 35 35 36 38 38 46 47 47 48 49 49 51 52 62 63 63 63 63 Gender M M F F M F M M M F F M M M F F M F TASK 1 – CAUSALITY # times knocked over 3/3 4/4 2/2 4/4 4/4 4/4 5/5 4/4 4/4 4/4 4/4 4/4 4/4 4/4 4/4 4/4 4/4 4/4 blocks / # of trials Average # of hits 1.3 1.0 1.0 11.8 1.5 1.8 1.0 1.0 1.8 1.0 1.8 1.3 1.3 1.0 1.0 1.0 1.8 1.0 required for task TASK 2 – NEGATION # times stopped / # of 7/10 0/6 10/14 4/12 8/8 8/8 7/12 10/10 14/16 8/8 8/8 8/8 10/10 8/8 8/8 11/11 8/8 8/8 trials Average # of hits 1.4 n/a 1.6 6 1.5 1.8 1.4 1.7 1.4 1.3 1.5 1.1 3 1.4 1.1 1.6 1.3 1.6 required for task TASK 3A – BINARY CHOICE # times turn appropri7/11 7/13 8/13 7/15 10/10 8/8 9/14 12/12 8/13 9/9 7/10 12/12 7/9 8/8 9/9 9/9 9/9 8/8 ately / # of trials TASK 3B – SEQUENCING # times knocked over 3/11 0/13 1/12 0/15 0/10 0/8 3/15 8/12 8/13 8/9 5/10 11/12 6/10 8/8 7/9 8/9 8/9 8/8 desired stack of blocks / # of opportunities LEARNING PROCESS FOR TASK 3 # of trials before success 2 n/a n/a n/a n/a n/a n/a 1 2 1 n/a 0 0 0 2 1 0 0 – Session 1 # of trials before success 0 n/a 2 n/a n/a n/a 1 0 0 0 0 0 1 0 0 0 1 0 – Session 2 Table 3 Pairwise comparison between mean success rates in different age groups – Welch’s tests p values Welch’s tests p values

Task 2 – Negation Task 3A – Binary Logic Task 3B – Sequencing

4 yrs old mean success rate > 3 yrs old mean success rate 0.044 0.063 0.002

5 yrs old mean success rate > 4 yrs old mean success rate 0.12 0.019 0.007

formance improved. The average number of successes in Task 2 for the four year olds was significantly greater than for the three year olds (Welch’s test, p = 0.044). The five year olds succeeded in all trials and their average number of successes was not significantly greater than for the four year olds (Welch’s test, p = 0.120). For Task 3A turning the wrong way was recorded as unsuccessful. Task 3B was recorded as successful if the child knocked over the blocks, even if the child used a different strategy than “turn first then go forward” with only two switch activations. Comparison of the average number of successes between the four and five years old groups and between the three and four year olds revealed that the five year olds performed significantly better in Task 3A than the four year olds (Welch’s test, p = 0.019), and that the four year olds

performed better than the three year olds, although the latter was not significant (Welch’s test, p = 0.063). In Task 3B, the average number of successes for the five year olds was significantly greater than for the four year olds (Welch’s test, p = 0.007), and this in turn was significantly greater than the average number of successes for the three year olds (Welch’s test, p = 0.002). The percentage of incorrect responses to the questions regarding the function of the switches are summarized in Table 4. Children aged three had more difficulty understanding the function of the Forward switch when the robot was facing the stack of blocks than the four year olds (40% of the three year olds gave wrong answers whereas only about 20% of the four year olds did). Three and four year old participants had problems in predicting where the robot would move if the Forward switch was hit when the robot was turned 90 degrees to the left (approximately half of the three and four year olds gave wrong answers) or when the robot was facing them (approximately 30%) gave wrong answers. Five year olds had no problem understanding the Forward switch function. The majority (70%) of the younger participants and approximately half of the four year olds (57% for the left turn switch and 37% for the right turn switch) were not able to correctly explain the function of the turn switches; 20% of the five year old children answered the questions regarding the


123

Table 4 Percentage of incorrect answers to the questions about the functions of the switches Question

% of incorrect answers 3 yrs 4 yrs 5 yrs 40 19 0 43 53 0 33 30 0 70 57 20 70 37 20 100 43 11

“When this switch [F] is touched, where does the truck go?” “If the truck is turned [90 degrees to the left] and I touch this switch [F], where will the truck go?” “If the truck is turned toward you [facing the child] and I touch this switch [F], where will the truck go?” “When this switch [] is touched, where does the truck go?” “If the wire to the switch is cut and I touch this switch, what will the truck do?”

Table 5 Technology survey results. (Frequency scores: 1 - Never, 2 - Weekly, 3 - Daily; Skill level scores: 1 – Low (trial and error), 2 – Medium, 3 – High (mastered); N/A: not applicable) Participant # Age (months) Gender On/Off switches Frequency Skill Level Proficiency measure

8 35 M 3 3 3

12 35 M 3 3 3

9 36 F 3 3 3

16 38 F 3 3 3

7 38 M 3 3 3

6 46 F 3 3 3

14 47 M 3 3 3

10 47 M 3 3 3

17 48 M 2 3 2.5

3 49 F 3 3 3

15 49 F 3 3 3

5 51 M 3 3 3

13 52 M 3 3 3

20 62 M 3 3 3

11 63 F 3 3 3

4 63 F 3 3 3

18 63 M 3 3 3

19 63 F 3 3 3

Multi-button remote controls

2 2 2

1 N/A 1

1 N/A 1

1 N/A 1

3 2 2.5

1 N/A 1

3 3 3

2 2 2

3 3 3

2 2 2

2 2 2

1 N/A 1

3 3 3

3 1 2

2 3 2.5

3 3 3

3 2 2.5

1 N/A 1

Frequency Skill Level Proficiency measure

turn switches incorrectly. All three year olds thought that a disconnected switch would still make the robot move, while 43% of the four year olds gave the same answer. In the five year old group the percentage of wrong answers to this question dropped to 11%. Results from the technology survey are compiled in Table 5. For each type of control a measure of proficiency was computed simply by taking the average of the scores in frequency and skill level. With this measure, a child that used one type of control weekly (score 2) with a high skill level (score 3) has the same 2.5 proficiency value as another child that uses the same type of control daily (score 3) but only with medium skill level (score 2). All participants used daily and mastered on/off switches but not multi-button remote controls. Correlation factors between the proficiency measure in using multi-button remote controls (see Table 5) and results for Tasks 2, 3A and 3B were computed, all yielding positive values less than 0.348 (Table 6). Therefore, it can be argued that the performance in the study tasks is not linearly dependent on previous experience in using multi-button remote controls.

4. Discussion All participants appeared to enjoy playing with the robot. However, five among the eighteen children were

Table 6 Pearson linear correlation factor between the proficiency measure in using multi-button remote controls and different task’s results Pearson linear correlation factor Multi-button remote control proficiency

Task 2 – Negation 0.348

Task 3A – Binary Logic 0.121

Task 3B – Sequencing 0.267

shy and did not want to enter the room for the first session and the researcher had to show them the robot in the hallway to convince them. For one of the participants it was necessary to have his older sister with him for encouragement. Once she played with the robot he performed the tasks and enjoyed playing with the roverbot. Two children required prompting to touch the switch; others started hitting the available switch immediately. All but one of the participants were comfortable with the roverbot by the second session. The results in Tables 2 and 3 show that proficiency in the tasks increases with age, as expected. All of the participants demonstrated skill in the first task, causality. Most of the participants hit the switch once to see what happened and then kept pressing it until the roverbot reached the stack of blocks and knocked it over. One participant (one of the two youngest) did not understand that holding the switch down would make the robot continue moving so she kept hitting and releasing the switch until the robot knocked over the stack of blocks (this participant hit the switch an average of 11.8 times

124


to accomplish the task). Forman [8] found that cause and effect skills varied across three year olds, whereas Stanger and Cook [20] found that two and three year old children consistently demonstrated causality. Negation, Task 2, had more mixed results, since this task was more difficult than causality for children aged three and four. The average number of switch hits to complete the task was always greater than one showing that every child refined the stopping position trying to get closer to the specified location at least once. Four year olds performed better than the three year olds. Five year olds completed the task in 100% of the trials. These results are consistent with Forman [8] who found that three and four year olds recognized that holding down a switch would make the robot move, but did not understand that releasing the switch (negation) is also a command (required to stop the robot), while five and six year olds had mastered this concept. In Task 3A, binary logic, even the youngest of our participants succeeded on most trials. This is in contrast to Forman where only children older than four demonstrated the binary logic concept. However, Forman’s study used one rocker switch with two directions of movement whereas this study used two separate switches for each direction located spatially on the left and right side of the forward switch. This additional spatial cue may have led to greater success. Again, five year old children succeeded in all trials. For Task 3B, most of the participants understood that to knock over one of the off-centre stacks of blocks it would be necessary to use more than one switch. In general, children aged four and five years old quickly understood this requirement. However, younger children often hit the turn switch several times, making the robot turn in circles, before understanding that the forward switch had to be hit to move the robot toward the stack of blocks after the robot was properly oriented. Other participants, having hit the turn switch a second time and acknowledging the error, purposely made the robot turn 360 degrees to return to the initial position. Then, starting over, they were able to “turn first then go forward”. Some of the older participants completed the task using alternative sequences of switch hits than just pressing turn and then forward. Participant #13, aged four, used sequences of left, right and forward hits to move the robot forward to knock over the blocks. Participant #5 knocked over the stack of blocks three times by hitting the left and right switches in sequence, causing the roverbot to move forward in a zig zag pattern. In some cases, multiple switch hits resulted from the way in which the child executed the task. Participant #10,

for example, hit the forward switch briefly in five of the trials before turning and moving forward again, always knocking over the desired stack of blocks. All children demonstrated some success at Task 3A. A number of children did not have any success at Task 3B. Some of the younger participants reoriented the switches so the arrow on the switch pointed in the desired direction of movement in an attempt to change the robot’s direction of motion. The number of trials before success in Task 3B diminished from session 1 to session 2, showing that children hold in memory what they learned from the previous session. In Stanger and Cook’s study, the three year olds could complete a two step sequence, but not three steps [20]. When the participants were asked about the functions of the switches the majority indicated that the forward switch made the robot move forward when the roverbot was pointed forward. Some of them didn’t understand that if the robot is pointing left or toward the child, the same switch will move the roverbot forward relative to its orientation, i.e. towards the left or towards them. They insisted that the roverbot would move forward with respect to their own position. One child said that the robot would drive towards him but that the forward switch would have to be rotated so the arrow faced him. The participants gave several explanations for the left and right turn switch function: i) the robot turns left or turns right (the correct answer); ii) the robot goes left or right (turns and moves forward in that direction); iii) the robot goes to the position where the stack of blocks was placed (they linked the actual function of the switch with the usage they made of it). Some of the participants succeeded in Task 3B even though they could not accurately describe the function of the switches. These erroneous explanations, along with the belief of younger children that a disconnected switch will still make the robot move and that by reorienting the switch the robot would move in another direction, are consistent with the results by Forman [8], where younger children believed that the action was in the switch, not in the relationship between the switch and robot. The absence of a high linear correlation between child’s proficiency in using multi-button remote controls and their performance in Tasks 2,3A and 3B shows that the results here presented were not biased by the children’s previous experience with switches. A limitation of the study is that the robot tasks were developed “intuitively”, with the expectation that they test the cognitive skills proposed. They have not undergone construct validity testing. There are standardized


tests for school age children, but they assume that fundamental skills such as these are already in place, since they usually occur before age 3 or 4 in most children. Sequencing is addressed and is a later skill closer to 4–5 years.

5. Conclusions This study provides data regarding the ages at which typically developing children demonstrate understanding of causality, negation, binary logic, and sequencing while using switches to control Lego robots. These data provide a means for estimating the cognitive developmental level of children with disabilities engaged in similar robot-related tasks. Establishing the level of understanding of these skills provides the opportunity to use the robot tasks as probes of cognitive understanding by children with disabilities. The robot task motor requirements are minimal and can be adapted to a wide range of possible anatomical control sites for activating the switch(es) (e.g., hand, head, leg, arm, etc.). There is also no need for spoken language to evaluate understanding. This is in contrast to children being underestimated due to the limitations of standardized testing procedures. One outcome that has been consistent in all of our robot studies is that teachers underestimated the abilities of the children until they saw their capabilities with the robot tasks [4]. The information gathered from typically developing young children using robots in this study and that of Forman can assist in establishing tasks that are developmentally cognitively appropriate which provide a challenge to the children and encourage development. (e g. [6]). The skills that were evaluated in this study have direct applicability to assistive technology use on a broad scale. Means end causality is a fundamental requirement for use of any switch activated electronic assistive device whether for simple appliance or toy activation or more complex alternative access methods to computers, environmental control units (ECU), powered mobility and augmentative and alternative communication (AAC) devices. Negation underlies the understanding that releasing a switch is an action that causes an effect. One example is inverse scanning used in AAC devices. In this mode, the cursor moves through selection elements until the switch is released at which time the selection is entered into the device [5]. This type of scanning is also used in mouse emulation for computers, menu control for ECU’s as well as other electronic assistive device applications. An understanding of binary

125

relations is necessary for driving a powered wheelchair with left and right capability. It is also important in the use of directed scanning in computers, ECU or AAC when using an on-screen keyboard. Finally, sequencing is a basic skill required in the use of computers, ECU or AAC for navigating the pages of an interface or to string together selections into meaningful commands or words. Given the importance of these skills for effective use of assistive technologies, it is important that there be meaningful assessment of these skills in children with disabilities. For many of these children assistive technologies are being considered because of lack of speech and/or severely limited motor skill. We have identified the cognitive skills relevant to the use of assistive technology, by using robot tasks which have low motor and linguistic demands. Hence, the robot tasks could be symbol and device independent ways of looking at very specific cognitive skills without the choice of a communication element, an environmental control function or a wheelchair direction causing additional cognitive overhead. The robot tasks could provide an opportunity for children to develop skills for more sophisticated assistive technology use, for example, beyond simple cause and effect computer games. The independence from motor or speech requirements of the robot tasks allowed us to use the tasks in a study with children who had severe disabilities and determine their levels of cognitive understanding when they were judged “untestable” by other standard measures [6]. Thus, robotic tasks such as those described in this study can be valuable in future studies as a proxy measure of disabled children’s cognitive ability.

References [1]

[2]

[3]

[4]

A. Andreopoulos and J.K. Tsotsos, A framework for door localization and door opening using a robotic wheelchair for people living with mobility impairments, Proceedings of the Robotics Science and Systems (RSS) 2007 Manipulation Workshop: Sensing and Adapting to the Real World, June 30, 2007, Atlanta. S. Besio, ed., Analysis of critical factors involved in using interactive robots for education and therapy of children with disabilities, Editrice UNI Service, Italy, 2008. A.M. Cook, P. Hoseit, M.L. Ka, R.Y. Lee and C.M. ZentenoSanchez, Using a robotic arm system to facilitate learning in very young disabled children, IEEE Transactions on Biomedical Engineering 35(2) (1988), 132–137. A.M. Cook, B. Bentz, N. Harbottle, C. Lynch and B. Miller, School-based use of a robotic arm system by children with disabilities, IEEE Transactions on Neural Systems and Rehabilitation Engineering 13(4) (2005), 452–460.

126 [5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

L. Poletz et al. / Robot skills and cognitive performance of preschool children A.M. Cook and J.M. Polgar, Cook & Hussey’s Assistive Technologies, Principles and Practice (3rd ed.), Elsevier Inc., Philadelphia, PA, 2008. A.M. Cook, K. Adams, J. Volden, N. Harbottle and C. Harbottle, Using Lego robots to estimate cognitive ability in children who have severe physical disabilities, Disability and Rehabilitation: Assistive Technology, in press. S.P. Eberhardt, J. Osborne and T. Rahman, Classroom evaluation of the Arlyn Arm Robotic Workstation, Assistive Technology 12(2) (2000), 132–143. G. Forman, Observations of young children solving problems with computers and robots, Journal of Research in Childhood Education 1(2) (1986), 60–73. W.S. Harwin, A. Ginige and R.D. Jackson, A robot workstation for use in education of the physically handicapped, IEEE Transactions on Biomedical Engineering 35(2) (1988), 127–131. R. Howell and K. Hay, Software-based access and control of robotic manipulators for severely physically disabled students, Journal of Artificial Intelligence in Education 1(1) (1989), 53–72. G. Karlan, S. Nof, N. Widmer, I. McEwen and B. Nail, eds., Preliminary clinical evaluation of a prototype Interactive Robotic Device (IRD-1), Proceedings of the ICAART 88, 1988, Montreal, Quebec. G. Kronreif, M. Kornfeld, B. Prazak, S. Mina and M. Fürst, Robot assistance in playful environment – user trials and results, Proceedings of the IEEE International Conference on Robotics and Automation, April 10–14, 2007, Rome, Italy. H. Kwee, J. Quaedackers, E. van de Bool, L. Theeuwen and L. Speth, Adapting the control of the MANUS manipulator for persons with cerebral palsy: an exploratory study, Technology & Disability 14(1) (2002), 31–42. H. Kwee, J. Quaedackers, E. van de Bool, L. Theeuwen and L. Speth, eds, POCUS project: adapting the control of the MANUS manipulator for persons with cerebral palsy, Proceedings of the International Conference on Rehabilitation

[15] [16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

Robotics (ICORR), July 1–2, 1999, Palo Alto, CA. C.R. Musselwhite, Adaptive Play for Special Needs Children, College-Hill Press, San Diego, CA, 1986. S. Nof, G. Karlan and N. Widmer, Development of a prototype Interactive Robotic Device for use by multiply handicapped children, Proceedings of the ICAART 88, 1988, Montreal, Quebec. M. Patrizia, M. Claudio, G. Leonardo and P. Alessandro, A robotic toy for children with special needs: from requirement to design, Proceedings of the 11th International IEEE Conference on Rehabilitation Robotics, 2009, pp. 918–923. B. Robins, E. Ferrari and K. Dautenhaun, Developing scenarios for robot assisted play, Proceedings of the 17th Annual International Symposium on Robot and Human Interactive Communication, 2008, pp. 180–186. J. Smith and M. Topping, The introduction of a robotic aid to drawing into a school for physically handicapped children: a case study, British Journal of Occupational Therapy 59(12) (1996), 565–569. C.A. Stanger and A.M. Cook, Using robotics to assist in determining cognitive age of very young children, Proceedings of the IEEE Conference on Engineering in Medicine and Biology, 1990, pp. 1911–1912. J.K. Tsotsos, G. Verghese, S. Dickinson, M. Jenkin, A. Jepson, E. Milios, F. Nuflo, S. Stevenson, M. Black, D. Metaxas, S. Culhane, Y. Ye, and R. Mann, PLAYBOT: a visually-guided robot to assist physically disabled children in play, Image & Vision Computing Journal, Special Issue on Vision for the Disabled 16 (1998), 275–292. I.C. Uzgiris and J.M. Hunt, Infant Performance and Experience:New Findings with the Ordinal Scales, Chicago: University of Illinois Press, 1975. WHO – World Health Organization, International Classification of Functioning, Disability and Health – Children and Youth version (ICF-CY), 2007, www3.who.int/icf/ onlinebrowser/-/icf.cfm?undefined&version=7.

99


“Enabling products”: Consumers with limited hand functions evaluate an automatic jar opener Sajay Arthanata,∗ , Vathsala I. Stoneb and Douglas J. Usiakc a

Department of Occupational Therapy, College of Health and Human Services, University of New Hampshire, Durham, NH, USA b Center on Knowledge Translation for Technology Transfer, School of Public Health and Health Professions, University at Buffalo, Buffalo, NY, USA c Western New York Independent Living, Inc., Buffalo, NY, USA

Abstract. Usability and accessibility are key design characteristics that enable consumers with disabilities to effectively and safely interact with mainstream products as they pursue independent living. This product usability study examines experiences of consumers with hand function limitations regarding the quality and value of a commercially available automatic jar opener. The product was designed for use by people with diverse hand functions and its development process included continuous consumer input. Findings from this longitudinal case study include high ratings of the jar opener on usability and other indicators of quality. Consumers’ reported value of the product remained consistently high throughout the study. Product use was consistent during the trial period evidencing high consumer satisfaction and product acceptance. This study attests to the potential benefits of involving consumers with disabilities in the development of products not only as a strategy to integrate inclusive design features, but also to broaden their market value. The paper also highlights the methodology and key concepts underlying the investigation of quality and value for products designed through this approach.

1. Introduction One of the major domains of day-to-day functioning involves the effective and efficient use of consumer products. In the mainstream market, people with functional limitations equally rely on consumer products to assist with everyday activities and independent living as other consumers. In designing products for optimal access and use, one must consider the sensory, perceptual, cognitive, psychosocial and motor demands placed upon the user, as well as the environmental conditions under which the device might be used.

∗ Address

for correspondence: Sajay Arthanat, Department of Occupational Therapy, College of Health and Human Services, 111 Hewitt Hall, University of New Hampshire, Durham, NH 03824, USA. E-mail: [email protected].

Taking into account the estimated 46 million people (15 years and older) with disabilities in the United States [21], along with the projected rise in elderly population from 36 million in 2003 to 72 million by 2030 [9], ensuring broad product accessibility and usability is not only critical for this vastly expanding consumer population, but also imperative from a commercial standpoint [3]. In this context, product design concepts such as Universal Design [4], Design-for-all (Design for all Foundation, n.d), and Trans-generational Design [13] have a common goal of facilitating the usability of products for all users regardless of their abilities or disabilities. Such innovative approaches to product development have implications for academic, business and consumer sectors. On the one hand, they broaden the market by including the niche along with the mainstream, but


100

S. Arthanat et al. / “Enabling products”: Consumers with limited hand functions evaluate an automatic jar opener

on the other, potentially pose a risk for an industry already successfully enjoying its mainstream market share. However, manufacturer’s market report following a year since the product’s launch, suggested that nearly a million units of the Lid’s OffTM automatic jar opener were sold [15], which evidences mainstream market penetration of the product and makes a case for applying universally designed features to a mainstream product to strengthen its original market share. All the same, will the use of universally designed features in mainstream products bring benefits to consumers with disabilities in addition to the mainstream population? The automatic jar opener presented and discussed in this paper offers a case in point to address these questions through a study of its efficacy. The Lids-OffTM automatic jar opener from Black & Decker was designed and developed through a consumer-centered process that integrated inclusive design concepts. This article reports the methodology and findings from a two-year longitudinal efficacy study of this product as tested by consumers with limitations in hand functions within their natural home environment. In doing so, the goal of this product efficacy case study is to demonstrate the impact of incorporating usability and accessibility features in consumer products; it will also introduce methodologists to a product evaluation framework and will make product designers and manufacturers aware of the potential returns of integrating consumer knowledge in product development.

2. Background By and large, opening vacuum-sealed jars poses significant difficulties for many consumers regardless of individual capabilities. The problem is compounded for children, the elderly and anyone with hand functions limited by weakness, pain, joint disorders, involuntary movements, deformities, and amputations. People who are unable to grasp, manipulate and twist the jar and its lid, or find it difficult or uncomfortable to do so, require assistance in food preparation or may even avoid the purchase of certain types of jars and its food contents. The Lids-OffTM by Black & Decker is an innovative consumer product designed to mechanically open vacuum-sealed and hard-to-open jars. This electrically operated device uses a unique, motor driven gear system that grips and breaks the vacuum seal on a jar to unscrew its lid. The Lids-OffTM is designed to work with a wide ranging variety of jars including those with metal and plastic lids. The product has a rotating turntable

Fig. 1. The Lids-OffTM jar opener by Black & Decker .

with two rubberized grips that adjust to different jar widths, up to 4.5 inches. On top, there is an upper housing with two grips that can be lifted up and lowered to accommodate jars up to a height of eight inches. For opening jars, the upper housing is lifted by the handle and the jar is placed on the turntable. The upper housing is then lowered down. When the device is activated, the rubberized grips lock the base of the jar and the grips on the upper housing rotate the lid counterclockwise to unscrew it from the jar. Upon completion of unscrewing of the lid, the upper housing is lifted up and the jar with the unscrewed lid is removed by the user, (Fig. 1). Support to its development came from the Rehabilitation Engineering Research Center on Technology Transfer (T2 RERC), located at the University at Buffalo. Operating under funding from the National Institute on Disability and Rehabilitation Research (NIDRR) since 1993, this center has helped product developers and other industry stakeholders bring Assistive Technology (AT) devices and mainstream products to market for the benefit of consumers with functional limitations. Well-proven product development theories such as Quality Function Deployment [1], User-oriented product development [14] and Consumer-idealized design [5] stress the importance of recognizing consumer needs and embracing consumer recommendations in the blueprint of products. Keeping this premise in mind, the development of the Black & Decker LidsOffTM involved systematic and rigorous consumer input across all its phases-from prototype to final product [11,12]. By means of several structured focus groups, consumers (both able-bodied and those with


101

Fig. 2. Operational framework of the product efficacy project.

hand function limitations) first acknowledged their needs and expectations with jar opening devices and products, and then rated their satisfaction with current options of opening jars. The ensuing information reflected the consumer need for a user-friendly jar opening appliance, and the market scope for pursuing its development. Subsequently, consumers listed, defined and prioritized the ideal functions and features required in an automatic jar. The Black & Decker design team consequently created three jar opener prototype models by integrating these functions and features. The models were then critiqued by the consumers to validate the inclusion of the prioritized functions and features including overall shape, button location, button size, button shape, type of handle, and type of lock/unlock activator. Based on the results, the design team made iterations to the prototype and incorporated additional design features into their final product version. The Black & Decker Lid’s OffTM automatic jar opener was launched in spring of 2003.

peting product as attested by key product usability indicators: effectiveness-the variety of jars that could be opened; efficiency-the time required to open jars; safety-prevention of spillage, breakage or damage of jars; and comfort-overall effort required by the user to open jars. Further, consumers who participated in the trial expressed their overwhelming satisfaction and desire to purchase the jar opener compared to the competing product [18,19]. The focus of this study was to investigate the ‘real world’ efficacy of the jar opener when used by consumers with limited hand functions in their home environment. Based on the long term experience of consumers with the product, the fundamental evaluative questions addressed here were: 1) What impact does the use of the jar opener have on the value of the device to the consumer? 2) What impact does the use of the jar opener have on the daily living of the consumers?

4. Methodology 3. Rationale The study is part of a broader research whose rationale was to validate the consumer driven approach involved in the development of the jar opener. This validation required documenting the extent to which the jar opener would improve the task of opening jars for consumers with functional impairments. An earlier efficacy study employed an experimental design to investigate the comparative quality of the jar opener against a competing product in a controlled (lab) environment [18,19]. The findings from that study revealed that the jar opener was superior to its com-

The framework of this product efficacy study, as illustrated in Fig. 2, was designed with the premise of investigating the consumer perceived merit and worth of the jar opener across different timeframes within a natural environment. According to product evaluation literature, the merit of a product refers to its intrinsic quality, while its worth refers to the extrinsic value attributed to it [16,18,20]. It is important to note here that these two constructs although independent, are not mutually exclusive for evaluation of product efficacy. In other words, “high quality” products, however meritorious in themselves, may or may not be considered “worthy” by users. On the other hand, no product can

102


be worthy without merit. Merit is necessary but not a sufficient condition for the goodness of a product. Therefore both merit and worth were considered as the founding constructs of this study. From a measurable or observable standpoint, a key indicator of product quality is its usability [2,7], which represents the product’s potential for success as the user interacts with it in a given context; while, a key indicator of the product’s value is considered to be its acceptance by the consumer within a given context. 4.1. Research objectives In accordance with the stated methodological framework, the specific objectives of the study were a) To identify and describe trends in consumer perceptions of the jar opener’s usability across six weeks of home trial from day one to the end of two-months (EOT); b) To assess consumer reported acceptance of the jar opener at the end of the 6 month study period; and c) To examine the impact of the jar opener on the consumer’s independence following two years of its usage. 4.2. Participants Fifty participants were recruited for the research through the Western New York Independent Living Project (WNY-ILP). They were recruited to participate in the overall product efficacy study, which consisted of onsite (lab-based) usability to compare Lids off with its competitor, followed by home trials, focus of this paper. Thus they represented a cohort that participated throughout the span of the product efficacy study, a determinant of sample size. The sample selection had multiple rationales. First, the size was considered appropriate in order to accommodate and maximize the variety of hand function impairments that limited an individual’s ability to open jars. Secondly, 50 participants were considered desirable for statistical comparisons at the earlier onsite trials. Third, the sample size was considered helpful in accommodating for potential attrition during the home trials, which addressed the use of the device and its reported long term impact such as independence and value. The participants were selected based on their need to open vacuum sealed jars and their limitations in hand functions that prevented them from opening jars independently. We point out that the jar opener was devel-

oped as a universally designed product, and had taken into account the perspectives of able-bodied consumers, with its initial sales reports confirming its success within the mainstream population. The goal of this research (as stated above) was to determine its usability and accessibility of this universally designed product for consumers with disabilities, and as such, the sampling focused on persons with disabilities. 4.3. Instruments In order to strengthen the validity of measurements, the study instruments were developed initially by task analysis of jar opening by consumers with functional limitations and then interviewing them regarding their difficulties, strategies and needs associated with the task. The analysis of interview data led to the identification of measurable indicators needed to account for the quality and value of the jar opener. These indicators were further validated by mapping them to a criteria matrix comprised of the seven Universal Design principles [4] and an adapted consumer-centered quality criteria [10,17]. Organizing the indicators in such a criteria matrix, as shown in Table 1, prior to their inclusion in the instrument ensured that they addressed the key quality, and value domains that reflect the efficacy of the product. The study instruments were derived from these indicators and designed in a way to capture consumer evaluations of the jar opener against its critical competitors [16] that is, the device alternatives they had been using or used in the past. A Day One Response Questionnaire was used to record the observations of the participants on the first day that the jar opener. This instrument primarily captured the consumer’s sequential reactions to the product starting with the instructions manual, the assistance involved in its use, ease of use, one-handed use, and size for storage. In addition, questions related to value attributes such as product aesthetics, personal affinity (“device for me or not?”) and expectations were also included. A Weekly Response Questionnaire was also provided to the participants to keep log of their usage of the jar opener over every week for six weeks. Using this questionnaire, participants reported the number and type of jars they opened using the automated jar opener. Any accidents involving spillage of the jar contents or breakage of the jar or lid was also indicated. Participant’s weekly perceptions of the jar opener’s usability and other aspects of quality were also captured in


103

Table 1 Indicator distribution by designer and consumer perspectives (lids off efficacy) 2

3

27

6

22

Other

8

3

3

8

3

2

Learnability

3

Affordability

12

8

Safety

27

Reliability

Flexible

Comfort

3

Reparability

18

Operability

Durability

Equitable

Portability

Effectiveness

PRINCIPLES**

UNIVERSAL DESIGN

T RERC CRITERIA FOR DEVICE EVALUATION

Total

57 99

Simple and 5

intuitive

1

1

7

Perceptive 4

information Error tolerant

4

2

9

4

8

22

5

14

5

1

35

1

36

Low physical effort Appropriate size and space

Total

55

3

12

13

2

1

88

11

41

16

3

20

1

10

10

254

*Lane and others, (1997); ** Center for Universal Design, (2002)

terms of its ease of use, effectiveness, reliability, safety, durability, and required storage space. A Mid-study (telephone) Interview was conducted following the two months to gauge participant’s interest in adopting the jar opener and purchase intent at that point. The actual value of the jar opener to the participant was tested at this point by asking this question: “If the study were to end today, would you purchase the jar opener from a portion of your monetary compensation?” Subsequently, an End-of-trial (EOT) Questionnaire was used at the completion of two months of home trial for participants to rate their overall perceptions of quality and value of the jar opener. At the end of the six month study, consumers were queried about their voluntary use (or abandonment) of the jar opener during the last four months of the trial when they kept no logs. Consumers were again given an option to purchase the jar opener using part of the compensation due to them. For those consumers who did purchase the jar opener, a two year follow-up (telephone) interview was conducted to verify the impact of the product on their independence and change in quality of life. All the above questionnaires urged consumers to comment on their thoughts concerning the quality and value of the jar opener as a reflection of their quantitative ratings. A list of the key consumer comments (both positive and negative) were compiled as findings.

4.4. Data collection The study protocol and the instruments were approved by the Social and Behavioral Institutional Review Board (SBSIRB) of University at Buffalo. The study consumers were provided with a brand new jar opener for the home trial. Along with it, a package consisting of instructions, study instruments and stamped envelopes for the survey responses were also provided. A consumer coordinator corresponded with the consumers to collect their Day One, Weekly and EOT responses and also conducted the mid-study, six month and the two year follow-up interviews over the telephone. 4.5. Data analysis Analysis of the jar opener’s home trial data was conducted systematically beginning with the day one data, followed by the weekly data and finally the EOT data. Comparisons of consumer ratings of usability and acceptance of the jar opener between these successive time points were conducted subsequently. Descriptive analyses of the day one, weekly, and EOT responses were first conducted exclusively. The frequency with which the participants rated the jar opener with respect to each quality and value indicator was calculated and plotted on a continuum ranging from

104

S. Arthanat et al. / “Enabling products”: Consumers with limited hand functions evaluate an automatic jar opener Table 2 Participant Demographics Age Mean

52.92

Gender MALE 10

Median

55

Female

40

Range

28-74

Total

50

Disability Arthritis Carpal Tunnel Syndrome Multiple Sclerosis Spinal Cord Injury Cerebral Palsy Fibromyalgia Muscular Dystrophy OTHERS

very low, low, somewhat high, high to very high rating. From the weekly responses, the total number of jars opened by all 50 participants for each week was computed and contrasted. Safety associated with use of the jar opener was computed by taking into account the total number of reported accidents in relation to the total number of jars opened in each week. The trends in consumer ratings of the jar opener was analyzed by comparing their mean values from week one to week six using repeated measures of ANOVA. The change in consumer ratings from Week one to Week six was determined using a paired t-test. The initial (perceived / potential) usability and acceptance of the jar opener reported by consumers on Day One and the final (actual) quality and value as revealed at the EOT were compared using the frequency with which the participants rated the jar opener at these two time points (Day One and EOT) along the five point scale. In addition, a comparison of the purchase intent (willingness to purchase the Lid-Off and the perceived cost) at the end of the onsite trial versus the purchase intent at the end of home trial (decision to purchase the Lid-off and the affordable cost) was made as a marker of the change in acceptance associated with the use of the jar opener. Data from the two-year follow up interview was analyzed and described using the breakdown of consumers in terms of how they perceived the usability and their level of acceptance of the jar opener along the scale points presented on the Likert scale. In addition, a Pearson analysis was conducted to ascertain the relation between jar opener’s usage and level of acceptance by consumer.

5. Results 5.1. Participant demographics As listed in Table 2, all 50 consumers were either adults or elders in the age range of 28 to 74. A majority

Frequency 12 7 7 5 4 3 2 10

Percent 24 14 14 10 8 6 4 20

(40 out of the 50 consumers) were females. All of them expressed difficulty in opening vacuum sealed jars attributable to wide ranging diagnosed impairments primarily arthritis, carpal tunnel syndrome, multiple sclerosis, and spinal cord injury that limited their ability to open jars. Figure 3a and 3b correspondingly illustrate the right and left hand function profile of the 50 participants. These figures list the percentage of participants reporting the nature and severity of hand function limitation associated with their inability or difficulty to open vacuum sealed jars. In general, a major portion of participants reported moderate to severe intensity of weakness, pain, fatigue and spasticity as contributing to their difficulty to independently open vacuum sealed jars. In terms of limitations specific to jar opening, only seven (14%) opened jars manually, while 17 (34%) were unable to open jars without assistance, as seen in Fig. 4. Among the consumers who opened jars using a chosen method, 23 (46%) opened jars by several ingenious strategies such as “tapping the jar’s top on the floor”, “running hot water on the jar” or “by twisting the jar’s lid while holding the jar between the knees”, and three (6%) consumers used a commercially available device for jar opening. 5.2. Day One evaluation of the jar opener by consumers The Day One findings are displayed in Fig. 5. On the first day of its use, nearly 95% of the consumers were satisfied with the instructions for using the jar opener, rating them as simple (20.8%)to very simple (75%) to follow. In terms of its appeal, about 85% of the participants perceived the jar opener to be attractive (35.4%) to very attractive (50%). As a reflection of its usability, approximately 96% of the participants were satisfied with their ability to use the jar opener who independently stated it to be high (16.3%) to very high (79.5%). Correspondingly, an equal proportion of participants (96%) indicated their satisfaction with


Fig. 3. (a) Sample distribution by right hand function limitations. (b) Sample distribution by left hand function limitations.

Open Manually-14%, Use a Device-6%

Use Own Method-46% Need Assistance-34%

Fig. 4. Distribution of consumers by their method of opening jars (N = 50).

105

106


Fig. 5. Day One evaluation of jar opener by consumers.

Fig. 6. Jar opener ratings on usability: Trend across initial six weeks.

the ease of use of the jar opener rating it high (19.1%) to very high (76.6%). About 62.5% of the consumers reported that the jar opener can be used with only one hand easily (25%) to very easily (37.5%). Close to 59% of the consumers found the compactness of the jar opener (the space required for storage) to be high (38.8%) to very high (20.4%). Overall, nearly 70%

of the consumers clearly perceived the jar opener as a device for me. 5.3. Weekly evaluation of the jar opener by consumers Table 3 reflects the jar opener’s effectiveness on the basis of the total number of jars opened in relation to


107

Table 3 Usage of the jar opener across six weeks Weeks Participants* (N) Total number of attempted jar opening by all consumers Total number of jars opened successfully by all consumers Avg number of jars opened by each consumer Range Median

1 49 194 191 3.82 0–21 3.00

2 48 155 152 3.04 0–15 2.50

3 47 129 126 2.52 0–15 2.00

4 50 164 161 3.22 0–20 2.00

5 50 134 132 2.64 0–20 2.00

6 47 154 151 3.02 0–20 2.00

* Note: A few consumers failed to send their responses for some weeks. Table 4 Safety associated with usage of the jar opener across six weeks Weeks

Type of Accidents

Number of jars opened Frequency of Accidents Cracked / Broke Lid Cracked / Broke Jar Spillage Dropped Jar Dropped Device

1

2

3

4

5

6

Totals

191 1 0 0 1 0 0

152 1 0 0 1 0 0

126 2 0 0 1 0 1

161 3 0 0 3 0 0

132 3 0 0 1 1 1

151 2 0 0 1 1 0

913 12 0 0 8 2 2

the total number of jars that needed to be opened by the consumers. Based on the report of jars opened by all consumers over six weeks, the automatic jar opener opened 913 jars out of the 930 jars that were attempted, a 98% success rate. The jars (2%) that could not be opened using the device included very large jars (with more than 50 oz of content), very small and thin jars, and bottles with thin caps. On an average, each consumer opened at least two to four jars every week with a maximum of 21 jars opened by one consumer in a particular week. Table 4 summarizes the safety associated with use of the jar opener computed on the basis of the reported accidents during its usage in relation to the total number of jars opened. In all, there were 12 accidents reported out of the 913 jars that were opened. Of the 12, eight were accounted for by spillage of jar content; there were two instances when the jars were dropped, and two accidents involved the jar opener being dropped from the work surface. Out of all the jars opened, there were no reported accidents of a cracked or broken jar or lid. For the results described hereafter, the 5 scale points on each item in the questionnaire were respectively interpreted as 1-Very Low, 2-Low, 3-Somewhat High, 4-High, 5-Very High. As seen in Fig. 6, the weekby-week consumer rating of the jar opener remained consistently high or very high on most of the usability variables, except for one-handed use which was rated between somewhat high and high. Table 5 presents the results of repeated measures ANOVA comparing over-

Percentage safety in relation to jars opened 98.69% 100.00% 100.00% 99.12% 99.78% 99.78%

all differences of quality and value perceptions. Alongside in the table, the differences between week one and week six ratings of the jar opener as computed by a paired t-test are also reported. In function of several missing responses and list wise deletion, the repeated measures ANOVA took into account a smaller sample as opposed to the entire 50 consumers. Although there were marginal differences from week to week in the reported usability of the jar opener across the six weeks, the repeated measures ANOVA indicated no significant changes in the pattern over this period. However, there were noticeable improvements in usability ratings from week one to week six on all usability variables except on one-handed use. The reported comfort involved in use of the jar opener had significantly improved (at the set alpha level of 0.01) from week one to week six. As shown in Fig. 7, the week-by-week consumer rating of the jar opener on quality variables associated with its design and construction showed a similar trend as its rating on usability variables. Consumers rated the jar opener’s portability, ease of storage (compactness), maintenance, durability, and appearance as high to very high (4 to 5 on the Likert scale) maintaining a consistently high trend across all six weeks. The consistency of these high end ratings, as shown in Table 6, was corroborated by the repeated measures ANOVA, which showed no significant differences in the week to week ratings. The paired t-test differences between week one and week six also did not indicate any significant differences (at set alpha level 0.01) for the same quality indicators. The maintenance requirement for

108


Table 5 Trends in quality (usability) ratings of the jar opener across six weeks Usability indicators Ease of Use

Comfort

Reliability

Access to controls

Independent use

One-handed use

Descriptive measures

Weeks

N Mean SD N Mean SD N Mean SD N

1 41.0 4.0 1.3 41.0 4.6 0.8 41.0 4.6 0.7 41.0

2 34.0 4.2 0.9 34.0 4.7 0.8 35.0 4.5 0.9 35.0

3 30.0 4.2 1.2 29.0 4.4 0.9 30.0 4.6 1.0 30.0

4 37.0 4.2 1.2 37.0 4.6 1.0 36.0 4.6 0.9 37.0

5 37.0 4.1 1.2 38.0 4.7 0.8 37.0 4.6 0.8 38.0

6 34.0 4.3 1.3 34.0 4.8 0.7 35.0 4.7 0.5 35.0

Mean SD N

4.7 0.6 42.0

4.7 0.6 36.0

4.6 0.7 30.0

4.6 0.9 37.0

4.8 0.5 38.0

Mean SD N

4.9 0.5 40.0

4.9 0.4 35.0

4.9 0.4 30.0

4.8 0.8 37.0

Mean SD

3.8 1.3

3.9 1.3

3.2 1.6

3.6 1.6

n# 14

Repeated measures ANOVA F Sig 1.02 0.41

Paired t-test week 1 & week 6 Sig. (2 tailed) 0.17 (N = 31)

14

2.14

0.07

0.01** (N = 31)

14

0.6

0.69

0.71 (N = 31)

15

1.25

0.07

0.57 (N = 32)

4.8 0.5 35.0

15

Not applicable*

1.0 (N = 32)

4.9 0.4 38.0

5.0 0.2 35.0

14

0.88

0.49

3.6 1.6

3.6 1.6

# Sample size for Repeated Measures ANOVA after list wise deletion. * Significance at .01 level. Repeated Measures ANOVA not meaningful because the mean rating for all the 6 weeks were 5.0 for the given n.

Fig. 7. Jar opener ratings on selected quality indicators: Trend over initial six weeks.

0.47 (N = 30)


109

Fig. 8. Value attributed to the jar opener: “Device for me or not?”.

the jar opener was only rated somewhat lower at week six as compared to week one at a p value of significance of 0.04. As seen in Fig. 8, the personal affinity or value associated with the jar opener as indicated by device for me or not was rated high to very high across all six weeks. A repeated measures ANOVA week-by-week comparison of the value rating was not possible as all the mean scores with the reduced sample size for the analysis (N = 14, following list wise deletion) were found to be 5.0. Although the paired t-test showed no significant differences, a noticeable increase in the value rating was observed between week one (Mean 4.6, SD 0.78) and week six (Mean 4.7, SD 0.65). 5.4. Comparison of Day One versus end-of-trial evaluation of the jar opener The comparative quality and value of the jar opener as rated by consumers at Day One and at end-of-trial (EOT) following two months of its use is displayed in Fig. 9. Almost equal proportions of consumers rated the ease of use of the jar opener as very high at Day One (77.8%) and at EOT (78.8%). At the same time, a larger percentage of consumers rated their ability to use the jar opener independently as very high at EOT (93.3%) when compared to Day One (79.6%). On the same note, compared to 37.5% of the consumers who rated one-handed use of the jar opener as very high at

Day One, 65.9% rated it to be very high at EOT. The appearance of the jar opener was rated as very high by almost 50% of the consumers both at Day One and at EOT. In contrast, the compactness of the jar opener in terms of the required storage space was perceived to be very high by 48.8% of the consumers at EOT, compared to the 20.4% at Day One. While 69.4% of the consumers rated their personal affinity to the jar opener as very high on Day One, 82.2% rated it as very high after using it for six weeks. In all, 47 consumers completed the six month long home trials of the product efficacy study, while three consumers discontinued from the study midway and were unable to be reached. The actual value attributed to the jar opener was gauged by comparing the purchase intent of these consumers following the onsite trial and the actual decision to purchase or return the jar opener at completion of the study after six months. Following its onsite testing, 47 out of the 50 consumers (94%) reported that they would purchase the jar opener as opposed to the competing product. Out of these 47 consumers, 37 (79%) actually purchased the jar opener using a portion of their monetary compensation, while one consumer who did not wish to purchase the product at the clinical trial in fact did purchase it at the completion of the study. The average perceived price of the jar opener at the clinical trial was $35.3 (SD. Dev 17.22, Median 30), while the average price the consumers were willing to pay for the product at the com-

110


Table 6 Trends in quality (usability) ratings of the jar opener across six weeks Usability indicators

Descriptive measures

Portability

N Mean SD N Mean SD N Mean SD N Mean SD N Mean SD

Maintenance

Durability

Appearance

Compactness

Weeks 1 42.0 4.3 1.1 42.0 4.9 0.4 42.0 4.6 0.7 40.0 4.4 0.6 41.0 4.1 1.1

2 35.0 4.3 1.0 35.0 4.7 0.6 35.0 4.6 0.7 35.0 4.4 0.7 35.0 4.1 1.0

3 30.0 4.4 1.1 30.0 4.8 0.5 29.0 4.6 0.6 30.0 4.4 0.6 29.0 4.4 0.7

4 37.0 4.5 0.9 37.0 4.8 0.8 37.0 4.6 0.8 37.0 4.4 0.9 37.0 4.2 1.1

5 38.0 4.4 1.0 38.0 4.7 0.6 38.0 4.7 0.5 38.0 4.4 0.8 38.0 4.3 0.9

6 34.0 4.3 1.1 35.0 4.8 0.5 35.0 4.7 0.6 35.0 4.4 0.8 35.0 4.3 1.0

Repeated measures ANOVA n# F Sig 15 0.77 0.57

Paired t-test week 1 & week 6 Sig. (2 tailed) 0.86 (N = 31)

15

0.37

0.87

0.04 (N = 32)

15

0.26

0.93

0.48 (N = 32)

15

0.83

0.53

0.60 (N = 31)

15

0.47

0.79

0.08 (N = 31)

# Sample size for Repeated Measures ANOVA after list wise deletion.

Fig. 9. Comparison of Day One versus end-of-trial evaluation of the jar opener.


111

Fig. 10. Consumer evaluation of quality and value of the jar opener after two years.

pletion of the study was $29.4 (SD.Dev 13.15, Median 30).

This number is consistent with the number who had preferred the competing product at the onsite trials.

5.5. Use/abandonment of the jar opener during last four months of home trials

5.6. Two-year follow-up evaluation of the jar opener

As shown by Table 7, forty seven (47) of the 50 who participated in the onsite trials did complete the home trials. Of these, 43 (or 91.5%) did use the jar opener in their home at least some of the time during the last four months of the study. Note that almost half of them (48.9%) or 23 people reported using it every time. With another 13 (or 27.7%) using it most of the time, we see about 75% as frequent users of the device. This part of the study did not require them to keep logs and so speaks to their voluntary choice for using / abandoning the device. Only four people used it rarely or very rarely, and gave reasons for their dissatisfaction.

Out of the 37 consumers who purchased the jar opener as part of the efficacy study, 33 again rated the quality and value attributes of product including its usability following two years of use. Four consumers were not available for the two-year follow-up interview. As seen in Fig. 10, nearly 22% of these consumers reported that they used the jar opener every time a jar had to be opened, while 30% indicated that they benefited from the device most of the time a jar had to be opened. As for the jar opener’s durability (quality), a majority of consumers (65%), reported that the device worked equally well for them as it did two years ago, and 11% rated that the device worked better for them than at the

112


Table 7 Consumer Use/Abandon of Lids-OffTM during last part of the study period How often did you use jar opener? Every time Most of the time Some of the time Rarely Very rarely Total that completed the study Dropouts Comments related to rare use Did not work well for me I really don’t know. Took it down and never put it back up Inconvenience with chord; disability status has Fluctuated; Getting help in opening jars.

Frequency 23 13 7 2 2 47 3

Percentage 48.9 27.7 14.9 4.3 4.3 100

1 1

2 2

1

2

beginning of the study. Further, a statistically significant relationship (p < 0.01) between usage frequency and reported independence in cooking (r = 0.62), and impact on quality of life (r = 0.72) was revealed by the Pearson correlation analysis.

6. Discussion The findings from the analyses of home trial data on the jar opener efficacy study demonstrate that the quality and value of the jar opener were rated high by consumers not only on the first day of usage, but also consistently throughout the trial period. The device was well utilized by all consumers to open a diverse variety of jars irrespective of their size, shape, content and type of lid. Only two percent of jars (19 out of 934 jars) could not be opened by the device. These few jars were exceptionally outside the parameters of the jar opener (i.e., too large to fit in the frame, too small to grasp, bottle caps not amenable to a twisting motion). The jar opener was remarkably safe given the large number of jars opened, the small number of reported mishaps, and the functional impairments of the users studied. In relation to their hand function limitations, consumers found the jar opener to be intuitive and easy to use independently. Notably, consumer ratings began high and remained relatively unchanged until the end of the first two months, so the rating trends were not sharply upward (from low to high). On the other hand, the overall comfort level involved in use of the jar opener had improved significantly from week one to week six. Thus, the apparent lack of increase in the quality and value ratings of the jar opener is not surprising, given that the ratings were high from the very beginning;

in fact, no change or decline in these ratings across the rest of the trial period speaks positively of user acceptance. There was noticeable contrast (increase) in ratings between Day One and EOT on some usability attributes such as in one-handed use and independent use of the device. On the same basis, between Day One and EOT, more consumers (49% as opposed to 20%) felt that the jar opener was highly compact to store in the kitchen. Most importantly, 82% of the consumers at EOT (in relation to 70% of the consumers at Day One) strongly agreed that the jar opener was a device for me. Based on the number of consumers who asserted that they would buy the jar opener as opposed to the competing product in the beginning, there was a 79% adoption or acceptance rate of the product by the end of the study. An analysis of the qualitative data from the EOT questionnaire collected after the six-week trial corroborates the above results and further explains the acceptance of the device by most participants. Table 8 summarizes what was perceived as most positive in the device. On the other hand, some of the participants who returned the jar opener at the end of the study commented on their perceived drawbacks in the product. Two participants who returned the jar opener reported that the “device did not work for me.” A few others provided reasons related to its incompatibility with very large and very small jars (“cannot open a large orange juice jar”). Some others had concerns with its size in accordance to their space in the kitchen. One of the common observations related to the jar opener’s portability. Since the power cord was short, the jar opener had to be set up very close to an electrical outlet. Interestingly, one participant said “I put it up and never brought it down to use it”. Tracking back the person’s responses through the weekly questionnaires revealed the person’s dissatisfaction at not being able to open a jar containing a favorite food item, which made the person not inclined to buy the device unless the motor was made powerful enough. Comments on the earlier mentioned EOT questionnaire captured details of several shortcomings expressed by both those for whom the device worked and for those it didn’t. Table 9 lists comments that express dissatisfaction with the device along with corresponding changes suggested by participants by way of device improvement. It should be noted that Black & Decker has brought out a newer version of this device as part of their family of jar opener devices, where they incorporated most of these changes.


113

Table 8 What participants liked most about Lids Off: End-of-Home-Trial Comments Features liked most by participant No comments Easy to use/operate – “Makes it easy to open jars that I couldn’t before”; “the ease of opening jars’;”easy to use”. “Less physical effort to open jars”, less painful – “pain free”; “Makes life less stressful”; Portable, lightweight – “It isn’t too heavy to carry”; “portable”; “lightweight design”; Reliable – “Consistent’; “works every time’; “effective”; Versatility –“Works well with more jars”; “adaptable to all types and sizes of jars”; “there is no limitation on jars; I can be sure no matter what size or type of jar, it will fit”. Stable; “stays put”; “The skid proof base the jars sit on”. “Attractive machine”; “appearance”; “looks nice in my kitchen’; “the look; neat looking on shelf”; “I have a little more independence using it myself’;“easier to prepare foods; This device helped me become more independent”; “don’t need any help to use it”; “Safe to operate”; “it holds the jar great so I don’t have to worry about dropping jar anymore”; “grabs the bottom of the jar; overhead internal locks holds the object in place throughout”; Size – “Does not take up space”; “compact”; “can leave it where handy”. Easy to maintain and to clean “Easily used with one hand or another”.

Frequency N = 50 6 21 3 5 8 6 4’ 4 3 4 3 4 1

Table 9 End-of-Home-Trial comments by participants about improvements in Lids Off Changes participant would like to see in the device

Frequency (n = 50)

No comments No change needed “Lock up position needs to be strong, doesn’t always hold. Difficult to lift up to locking position”;“sometimes hurts my elbow”; “It’s a little hard to lift the top to place the jar or remove it, but I always did it without any accidents”; “Would like to see the top (turning part) always at the top height so all you have to do is put jar underneath and push down. Much easier than trying to lift up” “More powerful motor so it opens a specific brand of maple syrup jar.” Color choice - availability in different colors.

15 5 3

Size and weight. Ex: “Needs to be smaller and lighter”; “it’s too large and cumbersome”. “Release button moved to a more neutral location for easier use with right hand. Lock (up position) or down pins need to be stronger, does not always lock into up position” Ability to open a wider range of jars- very large/tall jars. Ex: “juice bottles, gatorade, milkjugs, . . . ”; or very small jars – Ex: “Glue jars, steak sauces. . . ”; “Design a device to accommodate various sized objects”.

5

“Ability to work on metal canning jar lids”. Cord too short; Need long, retractable cord.

Benefit to consumers from the jar opener as reported after two years of its use, is evident in the continued use of the product and the temporal stability of consumer perceptions regarding the product’s quality and value, in particular its usability. A majority of the consumers continued to use the product frequently, while only a few chose to use it very rarely (13.5%) or discontinued using it (8%). The reasons for non-usage were mostly attributed to reduced need or available assistance to

Improvements to the newer version by the manufacturer

Addressed by the newer version. The motor is placed at the bottom rather than at the top, so lifting the top is easier.

1

Addressed. The newer version has a bigger motor.

2

Addressed. The newer version is available in white, black and silver colors. Addressed. Newer version is smaller.

1

Addressed. Newer version works by touch at any part of the top, instead of one button location.

5

The suggested jars with extreme specifications (very large or very small) are not addressed in the newer version, which continues with the original versatility as to the range of shapes, sizes and types of jars opened. See above Not addressed by the newer version.

1 4

open jars at home, as captured in comments such as “I really don’t use that many jars” and “Sometimes it is easier to ask someone in the household because it is faster”. On the other hand, nearly 65% of consumers perceived that the jar opener improved their independence in cooking, with 48.6% agreeing strongly. Almost an equal proportion (67%) agreed that the product made a positive impact to their quality of life. Most importantly, consumers who used the jar opener more

114


frequently tended to place a higher value on the jar opener. Those with the greatest functional need for an automatic jar opener valued it the most.

7. Conclusions The consumer home trials discussed in the paper are part of a systematic and comprehensive evaluation of the efficacy of the Black & Decker automatic jar opener in three phases. The participant consumers of the jar opener had previously attested to the usability of the jar opener when tested against a competing product during onsite trials, and they did so again when used in home trials (natural setting) over an extended period of time. The consumer ratings of the jar opener across all of usability and other key variables of quality remained consistently high throughout the trial period. The consumers accepted and valued the product. They continued to use the product without abandoning it beyond the initial two months of home trials even though they were not required by the study to use it and give feedback. Furthermore the majority of participants confirmed the product’s actual value by purchasing it at the end of the trial. Notably, consumers reported continued use as well as improved independence after two years, which indicates their satisfaction and acceptance of the product as well as a cumulative effect or impact of its use on their quality of life. The findings of the study support the validity of the underlying product development approach. This product was the direct outcome of a process that valued consumer input in the design and development of the jar opener and took into consideration the needs of a broad spectrum of consumers with potential hand function limitations. Consumers are crucial stakeholders in product development. The development process and the systematic evaluation of the transferred product, as documented and reported in the case of the jar opener support many of the theoretical perspectives about consumer centered product development. Further support needs to come from cumulative evidence from a series of similar case studies. The payoffs of involving consumers in product development, especially building their needs into the design phase, are significant in terms of consumer satisfaction, product acceptance and use. These then integrate into benefits in the form of improved quality of life for customers, and continued strong sales in the marketplace for the manufacturer. A further point to consider might be how the experience with this particular product shapes the consumer’s

perceptions of the manufacturer as a brand identity. How does the introduction of a highly useful and robust product shape the attitudes of consumers – and their network of family and friends – with respect to future purchase decisions, and in turn shape the practice of manufacturers? The answer might reveal a longer term payoff from this consumer oriented approach to new product development.

Acknowledgement This study was supported by funding from the National Institute on Disability and Rehabilitation Research (NIDRR) of the U.S. Department of Education (USDE) under grant number H133E030025. The opinions contained in this presentation are those of the authors and do not necessarily reflect those of the U.S. Department of Education.

References [1]

[2]

[3]

[4]

[5]

[6]

[7] [8]

[9]

[10]

Y. Akao, Quality Function Deployment: Integrating Customer Requirements into Product Design, Cambridge, MA: Productivity Press, 1990. S. Babbar, R. Behara and E. White, Mapping product usability, International Journal of Operations and Product Management 22(10) (2002), 1071–1089. S.M. Bauer and J.L. Lane, Convergence of assistive devices and mainstream products: Keys to university participation in research, development and commercialization, Technology and Disability 18(2) (2006), 67–78. Center for Universal Design, About Universal Design. Retrieved June 22, 2007 from College of Design, North Carolina State University Website: http://www.design.ncsu.edu/ cud/about ud/about ud.htm, 2007. S. Cincianntelli and J. Magdison, From experience: consumer idealized design: Involving consumers in the product development process, Journal of Product Innovation Management 10 (1993), 341–347. Design for all Foundation (n.d.), Design for all: What is? Available from http://www.designforall.org/en/downloads/ dossier-DfA-Fd-ang.pdf [Accessed 22 June, 2007]. W. Dzida, Standards for user-interfaces, Computer Standards & Interfaces 17(1) (1995), 89–97. S.H. Han, M.H. Yun, J. Kwahk and S.W. Hong, Usability of consumer electronic products, International Journal of Industrial Ergonomics 28(3–4) (2001), 143–151. W. He, M. Sengupta, V.A. Velkoff and K.A. DeBarros, 65+ in the United States: 2005. US Census Bureau – Current Population Reports: Special Reports, Washington DC: US Government Printing Office. Available from http://www. census.gov/prod/2006pubs/p23-209.pdf. [Accessed 26 April, 2007], 2005. J.P. Lane, D.J. Usiak, V.I. Stone and M.J. Scherer, The voice of the customer: Consumers define the ideal battery charger, Assistive Technology 9(2) (1997), 130–139.

S. Arthanat et al. / “Enabling products”: Consumers with limited hand functions evaluate an automatic jar opener [11] [12]

[13]

[14]

[15]

[16]

J.A. Leahy, Paths to market for supply push technology transfer, Journal of Technology Transfer 28(3/4) (2003), 305–317. J.A. Leahy, J.P. Lane and D.J. Usiak, Improving accessibility of new mainstream consumer products through participatory development, Proceedings from Annual Rehabilitation Engineering & Assistive Technology Society of North America Conference, Orlando, Florida, 2004. J.J. Pirkl, Transgenerational design: A design strategy whose time has arrived, Design Management Journal 2(4) (1991), 55–60. E. Rosenblad-Wallin, User-oriented product development applied to functional clothing, Applied Ergonomics 16 (1985), 279–287. H. Schulman, Letter to Shareholders from CEO Harry Schulman, Available from http://www.applicainc.com/media/ 2005%20Letter%20to%20Shareholders.pdf [Accessed 24 April, 2007], 2005. M. Scriven, Evaluation Thesaurus, (4th ed.), Newbury Park, CA: Sage.

[17]

[18]

[19]

[20]

[21]

115

V.I. Stone, S.M. Bauer, J.P. Lane, D.J. Usiak, Z. Khan and C. Prabhu, Wheelchair tie-downs: Ideal features and existing products, Technology and Disability 8(3) (1998), 159–178. V.I. Stone, D.J. Usiak and S. Arthanat, Assessing Efficacy of Assistive Technology Transfers: Validation of the T2RERC’s Technology Transfer Model, Paper presented at the joint conference of the Canadian Evaluation Society and the American Evaluation Association, Toronto, ON, 2005. V.I. Stone, M. Lockett, D.J. Usiak and S. Arthanat, Beyond Technology Transfer: Quality of Life Impacts from R and D Outcomes, Assistive Technology Outcomes and Benefits 6(1), in press. The Joint Committee on Standards for Educational Evaluation. The Program Evaluation Standards, (2nd ed.), Thousand Oaks, CA: Sage. United States Census Bureau (2002). Prevalence of types of disability among individuals 15 years and older: 2002. Available from http://www.census.gov/hhes/www/disability/ sipp/disab02/ds02t2.pdf [Accessed 26 April, 2007].