A Multi-User Integrated Platform for Supporting the ... - Science Direct

Available online at www.sciencedirect.com

ScienceDirect� Available online at www.sciencedirect.com

Transportation Research Procedia 00 (2017) 000–000

ScienceDirect

www.elsevier.com/locate/procedia

Transportation Research Procedia 27 (2017) 35–42 www.elsevier.com/locate/procedia

20th EURO Working Group on Transportation Meeting, EWGT 2017, 4-6 September 2017, Budapest, Hungary

A Multi-User Integrated Platform for Supporting the Design and Management of Urban Mobility Systems Tânia Fontes a, José Correia a, b, Jorge Pinho de Sousa a, b, Jorge Freire de Sousa a, b, *, Teresa Galvão a, b b

a INESC TEC, Campus da FEUP, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal Faculty of Engineering, University of Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal

Abstract Public transport networks were, in the past, mainly designed to maximize the efficiency of commuting trips. However, with such perspective there are considerable risks to marginalize some specific population groups (e.g. disabled, elderly, children, pregnant, people in poverty). For enhancing social inclusion and improving the accessibility of more vulnerable citizens, such networks are often redesigned and adjusted. Nevertheless, even with such adjustments, it is sometimes difficult to provide efficient services that fully address the real needs and capabilities of travelers, partially because of the failure in following the fast technological and demanding changes of modern societies. Taking in mind these challenges, we have developed a conceptual model to support knowledge sharing and decision-making in urban mobility, and to improve the way travel information is addressed. The multi-user integrated platform proposed in this work is supported by the idea that information from different channels must be centralized, organized, managed and properly distributed. This idea is grounded in two main principles: (i) past and real-time information from a wide range of sources is combined for knowledge extraction, and such knowledge is going to be used not only to allow travelers to better plan their trips, but also to help transport providers to develop services adapted to the needs and preferences of their customers; and (ii) information is provided in a personalized way taking into account socio-economical differences between groups of travelers. © 2017 The Authors. Published by Elsevier B.V. Peer-review under responsibility of the scientific committee of the 20th EURO Working Group on Transportation Meeting. Keywords: Social exclusion; urban mobility; public transport; multi-user platforms.

*

* Corresponding author. Tel.: +351-220-413-508 E-mail address: [email protected]

2214-241X © 2017 The Authors. Published by Elsevier B.V. Peer-review under responsibility of the scientific committee of the 20th EURO Working Group on Transportation Meeting.

2352-1465 © 2017 The Authors. Published by Elsevier B.V. Peer-review under responsibility of the scientific committee of the 20th EURO Working Group on Transportation Meeting. 10.1016/j.trpro.2017.12.158

36 2

Tânia Fontes et al. / Transportation Research Procedia 27 (2017) 35–42 Tânia Fontes, Vera Costa, Jorge Pinho de Sousa, Teresa Galvão / Transportation Research Procedia 00 (2017) 000–000

1. Introduction Social exclusion was a term coined by Lenoir (1974) who identified a set of groups encompassing mentally and physically handicapped, suicidal people, aged invalids, abused children, substance abusers, delinquents, single parents, multi-problem households, marginal, asocial persons, and others with social misfits. In the meantime, the literature has increased this list, and although there is a continuous debate on what social exclusion really is (Preston and Rajé, 2007), the initial concept tends now to be dynamic and relative (Schwanen et al., 2015). Several authors include in this definition individuals that are personally, financially or physically affected (Sen, 2000; Kenyon et al., 2002; Mackett and Thoreau, 2015). Kamruzzaman (2016) presents a set of several measures commonly used to quantify social exclusion. The use of these measures shows that some population groups behave differently and face specific challenges and problems in relation to transport and mobility, this meaning that transport plays a key role in the prevalence of social exclusion (Mackett and Thoreau, 2015). In recent years, transport and social exclusion have been addressed through a variety of EU projects (e.g. SocialCar (http://socialcar-project.eu/), the Cost Action TU1305 (http://www.tu1305.eu/), ACCESS2ALL (http://access-toall.eu/), or GOAL (http://www.goal-project.eu/)). These projects clearly show that transport-related social exclusion is more likely to affect some groups than others. Mackett and Thoreau (2015) explain that this exclusion is related with income, disability, age, gender and ethnicity. According to these authors there are barriers for traveling that affect some socially excluded people such as disabled (Rosenbloom, 2007), elderly (Titheridge et al., 2009), or people in poverty (Lucas et al., 2016), more than the rest of the population. Mackett and Thoreau (2015) explain that this can be related to cost, availability of transport, psychological and physical barriers, facilities, and information. A study conducted by Lucas (2012) shows that, since 2003, with the publication of a strategic plan in the UK, researchers, policy makers and practitioners from several countries became interested in adopting a social inclusion approach to transport planning. Since then, city planners started to remove or minimize different types of barriers, developing more conscious city plans. In the last decade, with the widespread of new technologies, problems related with information provision also began to be addressed. At this level, Kenyon et al. (2002) explain that the use of information and communication technologies could enable a new, virtual mobility, promoting an internet-based increase in accessibility, as an alternative to an increase in physical mobility. Moreover, information increasingly needs to be provided in a variety of formats, and to be displayed using a clear language and characters with large fonts. Mackett and Thoreau (2015) state that data should be provided in ways that take into account the characteristics of the population as a whole. The on-line information that shows routes should include a variety of accessible alternatives, and allow for a variety of walking speeds, since this may significantly affect the overall optimal route. Although new technologies have been consistently used to improve public transport, and new guidelines (e.g. Kenyon et al., 2002) and frameworks (e.g i-TRIP (Rajapaksha et al., 2017)) have been proposed, some key questions have not yet been properly addressed in the domain of information provision. In particular, we should be aware that: (i) despite the wide availability of mobility-related information, access and use of that information by social excluded groups is still very limited; (ii) the adaptation of mobility services to the particular needs and expectations of travelers, according to their specific profiles, is also rather limited in practice; (iii) the existing public transport models do not have enough flexibility and knowledge to follow technological and societal dynamics, namely to provide more customized services. To address the limitations and gaps identified in our literature review, travel data should be centralized, organized, managed and distributed, taking into account the particular needs of each citizen. Accordingly, this work proposes a conceptual model to frame a multi-user type integrated platform, supported by these principles, able to improve efficiency and inclusiveness of urban public transport systems. Such platform is based in the optimized use of mobility information. Section 2 of this paper presents the main concepts for designing a multi-user type integrated platform, able to address the identified limitations. Policy implications related with the development of this platform are discussed in section 3.


37 3

2. A multi-user integrated platform 2.1. Platform functionality and components To better address the mobility needs of travelers, information from different transport providers (e.g. buses, metro, ambulances, taxis) needs to be handled and made available in a central system. This system should aggregate and manage the information collected from several sources, including information provided by travelers, for example by collecting crowdsourcing information about a particular bus stop, vehicle or service. Such an approach will provide a holistic view of urban mobility that clearly expands traditional concepts and perspectives on public transport. Historical and real-time information, collected in the entire analysis environment, should be combined for knowledge extraction on travelers’ requirements and expectations. Such information will be used, both in a short and in a long-term perspective: (i) to re-design the services according to the citizens’ needs; and (ii) to improve the quality of information provided to travelers, authorities and decision-makers, thus enabling the development of innovative mobility services, business models and policies. In this process, data preprocessing is required in order to ensure data privacy. The following main concepts support the proposed platform: (i) centralization of information provided by public and private transport services; (ii) organization of that information in an open framework; (iii) collection and cocreation of knowledge; and (iv) redistribution of new, generated information, according to the specific needs of the different travelers’ profiles. Figure 1 shows the main components of the proposed platform.

Fig. 1. Components of the multi-user integrated platform.

38 4


2.2. General specifications From a technological point of view, the platform requires a high degree of standardization, at the different levels where data interchange and integration are key factors. Therefore, to guarantee interoperability within the system, it is necessary that the components “speak a common language”, through a defined protocol that establishes a common nomenclature, data types, message syntax, and encoding rules, following international standards. Security, redundancy performance, reliability, anonymity, and access control issues must be addressed and enforced through all components, namely: (i) in the data model central infrastructure; (ii) in the Application Programming Interfaces (API); and (iii) in the client applications (e.g. web and/or mobile applications). The necessary middleware must be developed to guarantee that third parties will be able to use or deploy services and contents. Various requirements for ‘technical openness’ (such as guaranteeing data is machine readable and available in bulk) should be considered, as a way to facilitate the integration of the different modules, and to allow the platform to be easily scalable and extended. 2.3. Data modelling and data management To create a reliable and robust collaborative system for urban transport, the stakeholders, the required data and the system's requirements must be carefully identified and characterized. This is crucial to increase the scalability of the system. Data modelling will be used to describe the information requirements, the type of information to be stored, and the structures and the relationships between those elements. For this purpose, three different types of data models will be developed, while progressing from requirements to the actual database of the information system:  a conceptual data model that includes a set of technological specifications about the data, and that will be used to discuss and tune initial requirements;  a logical data model, defining the data structures that will be implemented in the database; and  a physical data model, mapping the logical data model into a physical data model that organizes the data into tables, and accounts for access, performance and storage details (e.g. data type, the periodicity of data). With the fast recent technological developments, the collection and storage of big amounts of data, and the possibility to easily relate data from different sources, new opportunities for urban mobility have clearly emerged. In this context, mobility data is used to analyze the overall performance of the travel network, by identifying its particular gaps and weaknesses, and to better know the customers, in order to increase the quality of the service by performing minimal changes to the system. However, the massive nature and high dimensionality of big data have introduced unique computational and statistical challenges, including scalability and storage bottlenecks, noise accumulation, spurious correlations, incidental endogeneity, and measurement errors (Fan et al., 2014). Improving the way information is addressed and provided to travelers can significantly enhance the performance of such systems. For this purpose, new algorithms and other decision support tools should be developed based on the collected data, and KPIs need to be defined for the characterization of travelers’ profiles. Data mining and machine learning tools are used, along with multi-criteria decision-making approaches and multi-objective optimization. In addition, since these profiles can change along time, knowledge extraction methods from crowd-sourced (usergenerated) data, public transport providers’ internal information (e.g. historical and real-time information on the network, vehicles, ridership and ticketing), and external information (e.g. environmental, infrastructure, congestion, events) can be very useful. These procedures will allow the creation of new mobility services and business models, specifically tailored to the needs of traveler-trip profiles, thus minimizing the limitations of traditional systems. 2.4. Information provision and user interactions In the main facilities (stations, modal interfaces), information for supporting travel decisions is provided either by some customized services or, due to the high economic costs of these services, by physical maps and timetables. In the past, such graphical interfaces were restricted to static maps, usually available in stops and stations, but with the widespread use of electronic devices, a new paradigm was created. In fact, these new technologies are allowing the development of innovative solutions that significantly improve the quality of the information provided to travelers, including those from socially excluded groups. Most of these innovative solutions are based on two main ideas: (i)


39 5

information can be richer if the user interfaces are personalized; and (ii) the new generation of graphic interfaces can also provide feedbacks from passengers, hence promoting and fostering a collaborative environment between travelers and transport providers. Even with the widespread access to the internet and to personal mobile devices, it is important to acknowledge that some particular social groups may not have access to these technologies or may have low ICT literacy (e.g. for reasons such as age or educational level). Moreover, specific groups, such as blind people, have other specific needs. In all these cases, alternative ways are required to provide and collect information to and from these groups. Thus, the key information to develop personalized interfaces for each of the different traveler profiles must be carefully identified. On the other hand, different characteristics of traveler profiles may require different types of user interfaces (e.g. touch-based or voice-based). For example, children or disabled people may require interfaces that easily depict specific information, when and where needed. In certain situations, interactive interfaces are required but, in other circumstances, non-interactive interfaces are enough (such as an LCD display with no input from the user). Thus, interaction styles (e.g. instructing, conversing, manipulating, exploring) for each traveler profile must be analyzed and then, based on the results of such analysis, the most appropriate type of interfaces for each traveler profile will be selected. Nevertheless, one given traveler can fall into different profiles, along time. These variations can occur by a set of different reasons, such as age, trip role or purpose, and/or temporal or permanent disabilities. This means that the system must be capable of identifying such profile changes. Therefore, specific information must be collected and processed, to detect such variations and then update the (graphical) user interfaces accordingly. As referred above, besides providing information, the new systems are also used to collect feedback from passengers. Crowdsourcing and knowledge extraction methods can support these innovations, allowing people to have access to valuable information for travelling with higher comfort, efficiency and safety. City planners and mobility providers can also use the co-created knowledge made available by this platform, to redesign services in prioritized urban areas. Taking into account the previous arguments, the development of graphical user interfaces should encompass: (i) the analysis and selection of the interface type that is more appropriate to each user profile; (ii) the choice of the means to automatically update the traveler profile; (iii) the identification of the most useful information for each traveler profile; and (iv) the identification of the crowdsourcing and knowledge extraction methods to collect feedback from users. Based on this knowledge, a set of mock-ups for different groups of travelers is designed and experimentally evaluated, by adopting a user-centered development methodology. Such mock-ups must be assessed by a set of selected user groups, and the feedback of such evaluation process will be used to improve the quality of the interfaces and of the interactions with the users. In general, the organizational, technological and social innovations associated to these new procedures should be evaluated with the involvement of the relevant stakeholders, in order to check whether the proposed solutions respond as expected, and to refine those solutions as required. For this purpose, user interfaces and system procedures are being tested in real word conditions and considering travelers from different socially excluded groups. 3. Use cases and system scope Three main stakeholders (or actors) interact with the system proposed in this work: (i) the travellers, i.e. the users of public transport; (ii) the transport providers; and (iii) the authorities that regulate and/or supervise the different components of the urban space. Transport providers include a broad range of urban operators, from public companies run by the government, by a municipality or by specific authorities (e.g. responsible for buses, urban or sub-urban trains, or metro systems), to private companies (e.g. running taxis or buses), operating independently or under some kind of contract with, for instance, a given municipality. The level and frequency of interactions with the system depend on the type of actor and on his specific characteristics and requirements. Moreover, the possible actions / activities of a given actor, in terms of urban mobility, will possibly involve and be directly associated to interactions with the system. More concretely, we have that:  travellers can: (i) plan a trip; (ii) book a trip; (iii) make a trip; (iv) re-plan a trip; (v) change a trip; (vi) check past travel data; and (vii) check schedules, maps, alerts and general information;  transport providers can: (i) create new services; (ii) close existent services; (iii) re-plan or adjust current services; and (iv) issue and receive information (e.g. to and from citizens or authorities);

40 6


 authorities can have, once in a while, some interactions with the system, since they may, for instance, need to spread information on the use of public space, or issue and receive different types of alerts. In figure 2, a sequence diagram illustrates the interaction of these actors with the system. In this example, a citizen plans and makes a trip. The system first checks the traveller’s profile and, according to this analysis, returns the information that is required and adequate for him to plan a comfortable trip (this may be a multi-modal sequence of segments, with multiple connections and waiting periods). During the trip, an “alert” is issued by an authority (this may be, for example, information on a road accident). The system checks the alert relevance for this particular citizen and, if justified, the alert is sent to him. Then, the trip is re-planned, and a new schedule is produced to support, in real-time, the traveller’s decisions. In the meanwhile, transport providers and authorities are collecting and handling all alerts and messages, thus updating the current knowledge on the network and on its operational conditions. In this context, one of the most innovative features of the system is its capacity to handle information and respond according to the specific citizen profile. Travelling profiles are defined by a set of factors such as: (i) routines for public transport use (e.g. every day commuting movements, or once a week trips); (ii) knowledge of the transport network (e.g. a commuter, or a tourist); (iii) literacy level; (vi) social group; (v) mobility limitations (e.g. blind, pregnant, elderly); and (vi) age (e.g. children, elderly).

Fig. 2. Sequence diagram with an example of use (here, a citizen is planning and making a trip, while an alert is issued by some authority).

After classifying the user according to his profile, the system provides the right information taking into account the specifications of the travel request under analysis, through graphical interfaces designed for the user’s particular needs (e.g. touch-based or voice-based). Information used in the text messages is also adapted to that particular profile. Table 1 shows some examples of messages sent by the system (e.g. through a smartphone or LCD) to different profiles of citizens planning a trip. For some profiles, additional information needs to be added to the standard messages (such


41 7

information is here highlighted in italics). Thus, while for daily commuters the information can be provided in a synthetic way through standard messages (e.g. case 1, in Table 1), for citizens with ICT illiteracy, low knowledge of the system, and/or low knowledge of the city, more details need to be included (e.g. cases 2, 3, 4 and 5, in Table 1). In this model, to minimize the social exclusion and incentivize the use of public transport, discounts can be provided for some particular social groups (e.g. cases 2 and 5, in Table 1). 4. Conclusion In this work, we have developed a conceptual model (composed by a methodology and a multi-user integrated platform) to support knowledge sharing and decision-making in urban mobility, and to improve the way travel information is addressed. This conceptual model is directed towards the development of effective, efficient and affordable mobility alternatives, specially for vulnerable to exclusion groups. Still, as recommended by Thoreau (2015), such system also takes into account the requirements of the whole population. The system is supported by the idea that information from different channels should be centralized, organized, managed and properly distributed. Historical and real-time information, collected within the entire analysis environment, must be combined for knowledge extraction and used, both at short and long-term, to re-design the services according to citizens’ needs, and to improve the quality of information provided. This will support the development of innovative mobility services, business models and policies. The main advantages for each type of stakeholders are:  travellers have a general vision of the available transport services, as available in a centralized system – the information received is personalized according to a profile that is dynamically adapted (a specific traveller can change the trip plan of an on-going trip without the need of finishing the initial plan);  transport providers can adapt and/or create new services and business models, able to address the population needs, namely taking into account the promotion of accessibility and social equity;  authorities can enhance the management of public infrastructures, e.g. during emergency situations, and effective inclusion policies can also be designed. Acknowledgements The work of the first author is funded by Fundação para a Ciência e Tecnologia (FCT), Portugal, through grant SFRH/BPD/109426/2015. References Fan, J., Han, F., Liu, H. 2014. Challenges of Big Data Analysis. Natl Sci Rev. 1(2), 293–314. Kamruzzaman, M., Yigitcanlar, T., Yang, J., and Mohamed, M.A. 2016. Measures of Transport-Related Social Exclusion: A Critical Review of the Literature, Sustainability, 8, 696. Kenyon, S., Lyons, G., Rafferty, J., 2002. Transport and social exclusion: investigating the possibility of promoting inclusion through virtual mobility. Journal of Transport Geography, 10(3), 207–219. Lenoir, R. 1974. Les Exclus: Un français sur dix, Editions du Seuil, Paris. Lucas, K., Mattioli, G. Verlinghieri, E. and Guzman, A. 2016 Transport poverty and its adverse social consequences, Transport 169(6), 353-365. Mackett, R.L. and Thoreau, R. 2015. Transport, social exclusion and health, Journal of Transport & Health, 2(4), 610–617. Preston, J. and Rajé, F. 2007. Accessibility, mobility and transport-related social exclusion, Journal of Transport Geography, 15(3), 151–160. Rajapakshaac, P., Farahbakhsha, R., Nathanailb, E., Crespia, N. (2017). iTrip, a framework to enhance urban mobility by leveraging various data sources, Transportation Research Procedia 24C(2017), 113–122. Rosenbloom, S. 2007. Transportation Patterns and Problems of People with Disabilities In: The Future of Disability in America. Editors: Marilyn J. Field and Alan M. Jette, The National Academy Press, Washington D.C., pp 519-529. Schwanen, T., Lucas, K., Akyelken, N., Solsona, D.C., Carrasco, J.A., Neutens, T. 2015. Rethinking the links between social exclusion and transport disadvantage through the lens of social capital, Transportation Research Part A: Policy and Practice, 74, 123–135. Sen, A., 2000. Social Exclusion: Concept, application and scrutiny. Office of Environment and Social Development, Asian Development Bank. Titheridge, H. Achuthan, K., Mackett, R.L., Solomon, J., 2009. Assessing the extent of transport social exclusion among the elderly. Journal of Transport and Land Use, 2(2), 31-48. György, K., Attilab, A., Tamás, F. (2017). New framework for monitoring urban mobility in European cities. Transportation Research Procedia 24C(2017), 155–162.

W

W

M

3

4

5

70

42

35

74

Y

N

N

N

ITS illiteracy

435 First Av  Marie Hospital

Reduced High mobility comfort / ITS illiteracy

Tourist (nonnative speaker) Airport  Fashion Hotel

34, John Smith Rd  2500, Central Av

Central Market  Central Plaza

Low travel time / High comfort

Pregnant High comfort / Low travel time

Elderly

Central Market  Central Plaza

Origin / Destination

-

09:20 p.m.

04:30 p.m.

-

-

8

8:25 a.m. – catch metro line A in First Av” metro station (line A is located in the 3rd floor; elevators are in the right side of the main entrance; the second carriage is for priority citizens) 9:35 a.m. – get out at the “Central Plaza” station (take the elevators at your left) 9:40 a.m. – catch a paratransit taxi to “Marie Hospital Marie” 9:50 a.m. – you have arrived to your destination Travel price: 3.80 € (discount price – you have saved 2.3 €)

9:20 p.m. – go to the “Airport” metro station (the station is 150 m distance from airport exit B) 9:33 p.m. – validate your travel ticket in the entrance of the metro station 9:35 a.m. – catch metro line C (during your travel the vehicle has 8 stops) 10:15 a.m. – get out at the “National Art Museum” station (leave the station on your left side) 10.16 a.m. – in the entrance, turn left and walk straightforward along 400 m of the Central Av (~5 min) 10:21 a.m. – you have arrived to your destination Travel price: 4.50 €

4:30 p.m. – go to the “Central” railway station (the station is located 200 m south of 34 John Smith Rd) 4:40 p.m. – validate your ticket in the entrance of the railway station 4:50 p.m. – catch metro line A (go to the 1st floor; elevators are in the left side of the main entrance; the first carriage is for priority citizens; during your travel the vehicle will stop 4 times) 5:10 p.m. – get out at the “Green Park” station; use the south exit of the station 5:15 p.m. – catch a taxi for 2500 Central Av 5:20 p.m. – you have arrived to your destination Travel price: 2.90 €

8:10 a.m. – catch bus 34 at the “Central Market” stop 8:30 a.m. – get out at the “Water Centre” stop (leave the station on your right side) 8:40 a.m. – catch metro line B in “Water Centre” metro station. Validate your travel ticket in the entrance of the metro station. Take the elevators at your left to platform A 8:50 a.m. – you have arrived to your destination. Central Plaza is in front to you Travel price: 0.9 € (discount price – you have saved 0.9 €)

8:10 a.m. – catch bus 34 at the “Central Market” stop 8:30 a.m. – get out at the “Water Centre” stop 8:40 a.m. – catch metro line B in “Water Centre” metro station 8:50 a.m. – you have arrived to your destination (Central Plaza) Travel price: 1.80 €

2214-241X © 2017 The Authors. Published by Elsevier B.V. Peer-review under responsibility of the scientific committee of the 20th EURO Working Group on Transportation Meeting.

10:00 a.m.

-

-

09:00 a.m.

09:00 a.m.

Information provided by the system Time to Time to start the arrive in the travel destination

W – women; M – men; Y – Yes; N - No / italics: additional information provided to particular groups of citizen.

M

2

1

Preferences

Gender Age Frequent Other user (Y/N) charact. M 29 Y Low price

Case Profile

Table 1. Illustrative cases of system utilization.

Tânia Fontes, Vera Costa, Jorge Pinho de Sousa, Teresa Galvão / Transportation Research Procedia 00 (2017) 000–000

42 Tânia Fontes et al. / Transportation Research Procedia 27 (2017) 35–42