Evaluating contextual information for wearable ... - Semantic Scholar

14 downloads 39912 Views 1MB Size Report
The study involves WECAPC software developed to run on the χ3 ... wearable computer with simple context-awareness yields superior .... best of both worlds.

Evaluating contextual information for wearable computing Huw W. Bristow, Chris Baber, James Cross and Sandra Woolley University of Birmingham, Edgbaston, Birmingham. B15 2TT, UK [email protected]

Abstract Does the use of a wearable context-aware computer increase task proficiency? User trials have been carried out in an attempt to answer this question. The study involves WECAPC software developed to run on the χ3 wearable computer. The trials demonstrate that a wearable computer with simple context-awareness yields superior performance to either Internet search or realworld activity. Thus, we conclude that the use of context does indeed improve user task proficiency.



Context plays an important role in tasks involving object identification, learning and memory. For example, Palmer [1] asked participants to identify objects shown on cards. People consistently performed best when a picture demonstrating the context of the object was shown before the object itself. In these, and similar studies, the role of contextual cues is to ‘prime’ the recognition or recall of specific information. The notion is that information is stored (in human memory) in a set of associative structures that link pieces of information together, and that providing cueing information somehow accesses an appropriate area of the structure. In recent years computing systems have begun to use context to aid the user. Simple examples of this can been seen in the introduction of ‘smart’ menus, which offer the user a restricted choice of functions based on the most commonly used functions over the previous few interactions. This is the most basic form of context awareness and simply asks, “what are the preferences of the person doing the task?” Often such a ‘dumb’ version of context-awareness interferes with user performance, e.g., by removing desirable functions from the restricted set, or not realising that a commonly used function might reflect a very specific (rather than very common) set of actions. One reason why these approaches appear to be ‘dumb’ is that they fail to key into an associative structure that links functions to actions. Recent developments in wearable and mobile computing extend the idea of context to relate to concepts such as the user’s location and activity. Ostensibly this could allow

rudimentary associative structures (or ‘models’) that link features of context to items of information; as the features of context change, so too will the relevance of information. Thus, a significant aspect of wearable and mobile computing is the ability of the technology to respond to changes in context. Unfortunately, there remains a distinct lack of agreement as to what constitutes ‘context’. Abowd and Mynatt [2] suggest that one can think of context in terms of: Who is using the system; What the system is being used for; Where the system is being used; When the system is being used and Why the system is being used. This provides an initial avenue into the problem of defining context. We suggest a classification scheme illustrated in table one ([3]). Table one pairs a reference marker, i.e., the element that is being defined, with a simple demarcation of time. In this manner context could be defined in terms of a combination of Reference Markers that have different information relating to whether the information is stored or whether it is predicted, or whether it is being captured at the moment. Table One: Features of Context Reference Markers Past Current Event Stored in Incoming diary message

Future Reminder of meeting


Stored in maps




Previous actions

Current performance

To Do list


Stored photo or name

Physiological status

Medication reminder


Virtual messages

Status display

Maintenance call

Most of the examples are reasonably self-explanatory, but one or two might require explanation. Therefore, the entry of “virtual messages” refers to the notion of leaving a message at a location that can only be picked up by someone with the appropriate technology, e.g., one could leave a text message on the card-swipe of a secure door and when the appropriate person enters, they could collect the message. One could further combine these examples into new products, e.g., combining : incoming call; with : sitting in a meeting and : in the bosses office, could lead to a device that recognises what is happening, does not seek to disturb you and routes the call to an answerphone. As we shall see in section 2, many applications that illustrate context-aware computing tend to focus on pairs of the reference markers in their system. In the study in this paper, we consider and . The main question for this paper is whether context- aware computing significantly enhances user performance in information retrieval tasks, in comparison with other means of retrieving this information. To relate this the initial discussion, we can ask whether a context-aware computer system cues relevant information in a manner that supports users in searching and retrieving this information.


Context-Aware Wearable Computing

There has been a great deal of research combining wearable computers and context-aware applications. One of the earliest papers described ‘A Touring Machine’ [4]. In this application, environmental reference markers (using GPS), were used to link a person’s location with both pertinent information, e.g., a description of the building that they were near, and augmentation of their view of the world, i.e., through superimposing arrows on the head-up display to point to buildings. Such a system is computationally intensive, and the computer was ‘worn’ in a rucksack, together with a hand-held tablet. One of the problems with this system was the sheer size of the equipment. Using a HMD and tablet has the advantage of a versatile display but raises problems of dividing attention between displays and also means that the user must use both hands for interaction. The paper also lacked evaluation of the usefulness of the wearable system, focussing on the technical aspects. Context could be defined by environment, i.e., a device could be used to detect changes in ambient sound levels or wearer movement (Schmidt et al., [5] report such a device that is mounted in the wearer’s tie). There has been a great deal of interest in the notion of sensors mounted in badges that could be used to detect a person’s location, and can modify their environment accordingly.

For example, an office worker could walk into an office and, once their presence had been detected, all of that person’s telephone calls and email routed to the desk in this office ([6]). Alternatively, context could be defined by changes in the physical and physiological state of the wearer ([7]; [8]). Finally, context can be related to the task that the person is performing. Thus, wearable computers can assist the wearer in shopping activities, ([9]). Most research into contextual computer systems has concentrated developing computer systems that can be worn (or carried) and that can respond to one or two Reference Markers. What has not been addressed in any critical manner in whether context-aware wearable systems really help people to complete their given task more proficiently? In order to address this question, a user trial was run. Before describing the trial and presenting the results, the next sections describe previous reports of user trials with wearable computers, and the system that was used for the research.


User Trials of Wearable Computers

There are a few papers presenting performance data from field trials of wearable computer. Some papers report comparison of different forms of interaction device for wearable computers, e.g., Thomas et al. [10, 11]. Other researchers have compared the manner in which users of wearable computers interact with information. Lehikoinen [12] investigated using wearable computers to host digital maps and has evaluated their system, again on a university campus. The research considered different ways of presenting maps to users of wearable computers, i.e., by automatic updating or by manual updating. However, the study did not include a control group, i.e., it did not investigate whether their system was an improvement on a conventional map, which may have been useful. Finally, several papers have compared information displayed on HMD connected to a wearable computer with a paper version of the same information. Thus, Siegel and Bauer [13] suggested that inspection took longer with the computer (28 minutes to complete all tasks using paper vs. 42 minutes using the computer). The authors suggest that the computer required users to reposition the display to either read instructions or to compensate for glare from the sun (the trials were conducted out of doors). Baber et al. [14] found that time to enter patient vital signs by paramedics were 97.4s without the computer, and 114.1s with the computer. Again, much of the variation in performance could be attributed to the need to adjust the HMD. In a study of aircraft inspection (with and without wearable computer), Ockerman and Pritchett [15] found no significant difference between fault detection checks, but significant differences in the manner in which inspections were

performed. “The wearable computer users jumped from one item to the next as the computer told them to do, whereas the pilots who did not use the wearable computer had continuous activity.” [p.37]. A similar finding was also reported in the Baber et al. [14] study. One of the problems with comparing wearable computers with paper manuals is the lack of obvious similarity between conditions. A more logical comparison would be to contrast information from two different computers, e.g., wearable vs. desktop, in the performance on tasks. In the trials presented in this paper, the aim has been to compare three conditions that reflect realistic sources of information for the task ‘find out about buildings on the University of Birmingham campus’. One source of information is the internet; people can access web-pages to find out about departments and buildings. A second source is the buildings themselves; people can walk around campus and look for information. A wearable computer ought to permit access to the internet for people who are walking around campus and should provide the best of both worlds. In terms of reference markers, the study uses , i.e., GPS data, and , i.e., find out about specific building.

3 3.1

The χ3

load in laboratory settings). While these times are, perhaps, too short for commercial application (one would not want to keep running out of power towards the end of the working day), they do suggest that it is possible to control power management in windows to an acceptable degree.

Figure 1 - The χ3 It can be seen from Figure 1 that the χ3 is quite small, measuring only 170 mm by 40mm by 100mm. Even with the addition of the head mounted display and battery the system is still comfortable, light and easy to wear. The entire system is housed in a shoulder strap, as shown in figure 2.

Technical Setup

The χ uses a PC104 embedded PC board (for more information see www.wear-it.net). It has SVGA out, two communications ports, on-board LAN and USB sound. In the version used in this study, the main unit is a 166Mhz Pentium class chipset (although recent versions have gone up to 700MHz). A MicroOptical head-mounted display is used (with its own power source and data converter), with the addition of a SVGA to NTSC converter allowing the screen to be made larger for reading text from webpages. A Garmin Global Positioning System (GPS) is used for tracking the users location. The GPS is accurate to a few metres, although can be affected by reflections from buildings (see Discussion). Whilst not used in the trial, a USB sound card provides access to speech recognition software and a GSM mobile phone supports mobile connection to the Internet. 3

The processor runs Windows95. This offers the capability to run commercially available software and to share files between different computers easily. However, there is the assumption that windows is ‘power hungry’. We have found, through work carried out the third author, that it is possible to modify BIOS settings in such a way as to significantly reduce power requirements and to extend battery life; we typically expect some 6-8 hours of battery life (and have managed to run for 10 hours on full

Figure 2 - The first author wearing the χ3



WECA-PC is the software that analyses the data from the Communications Port and sends a URL from a database to Internet Explorer. The incoming data either arrives from a GPS or from infra-red transmitters located in buildings (although the latter option is not used in this trial). The software is written in Visual Basic. The

software, as shown in figure 3, strips longitude and latitude coordinates from the GPS data stream. In addition to extracting coordinates, the software also performs error checking and invokes different routines depending of the absence or weakness of signals from the GPS. The coordinates are used to query a database of previously identified locations in order to call up a URL. The URL is then passed to Internet Explorer to search the World Wide Web.

average time, for the mobile conditions, taken walking between building and for the Internet condition the average time rested between questions); and that the internet would lead to quick information retrieval, and so should support significantly faster time to find information (particularly when one considers that the internet condition would employ the traditional mouse and keyboard and a 17” monitor familiar to all participants).



A set of questions about buildings around the university campus was devised. A set of questions refers to the two questions for each building. The questions could be answered using information in the real world, information from the Internet or information from either source. All questions were answerable by all conditions (Although efforts were made to make each question equally easy to answer using each source, the users generally found one source easier than another). 27 undergraduate students participated in the study (23 male and 4 female). Participants were then allocated, on appearance at the laboratory, to one of three conditions:

Figure 3 - System Architecture


User Trial

The aim of the user trial was to determine whether a context-aware wearable computer would be useful for information retrieval tasks. It was assumed that such a device would allow users to extract information from both the World Wide Web and from the world around them. Consequently, the trial was constructed so that people would search for information in one of three conditions (web only, world only, wearable computer). The comparison, between world, web and wearable, was intended to test whether wearing a context aware device led to superior performance to simply using information in the world, and whether mobile access to the web led to superior performance to simply accessing the web on a static terminal. We assumed that the two reference conditions, i.e., walking around and using the internet, would produce superior performance on two measures, i.e., we hypothesised that wearing the computer might impede mobility, so that the walking around condition should lead to significantly faster mean time between questions (note the mean time between questions is the

Condition 1: Wearable Computer - The user wore WECA PC and therefore had access to both environmental and virtual data. As the users approached the relevant building a single web page was displayed, in each case this page contained the answer to at least one of the questions asked about that building. Condition 2: World Wide Web - involved the user only having access to only virtual data. The users were asked to sit at an internet ready terminal and only use the University of Birmingham web sites to find the answers. Condition 3 World only - involved only environmental data. The users were asked to walk around campus with a paper map, visiting the relevant building to answer the questions. Each group was given a brief explanation of the task and an introduction to the equipment used. Training was limited as we wanted to see how people would cope with the novelty of the wearable computer. The questions were given to participants, and any queries regarding task or questions addressed. Each group was then told that they had twenty minutes to complete the task (although they were not stopped if they went over this time). The time taken to answer each set of questions was recorded (i.e. the time taken looking for information). This allowed comparison to be made of the time taken to complete each individual question. Users were asked to record whether they answered the questions making use of virtual or environmental information and any problems

they encountered. At the end of the experiment, the answers were checked and the participants debriefed. A full set of results was made available to participants the week after the experiment.



The results are divided into two parts: the first part considers the effect of the different conditions on performance time, and the second part considers the effect of the different conditions on participants’ ability to answer the questions.


Kruskal-Wallis tests were carried out on the time between questions and the total time taken to complete the task. For the ‘Time between Questions’ [χ2 = 6.11; df = 2; p

Suggest Documents