W3C homepageWeb Accessibility initiative homepage

WAI: Strategies, guidelines, resources to make the Web accessible to people with disabilities

3 December 2014

WAI R&D Symposia » Way-Finding Home » Proceedings » This paper.

This paper is a contribution to the Accessible Way-Finding Using Web Technologies. It was not developed by the W3C Web Accessibility Initiative (WAI) and does not necessarily represent the consensus view of W3C staff, participants, or members.

Extended Abstract for the RDWG Symposium on Accessible Way-Finding Using Web Technologies

mPASS: Combining Crowdsourcing and Sensing for Pedestrian Accessible Way-Finding

  • Silvia Mirri. Department of Computer Science and Engineering - University of Bologna - silvia.mirri@unibo.it
  • Catia Prandi. Department of Computer Science and Engineering - University of Bologna - catia.prandi2@unibo.it
  • Paola Salomoni. Department of Computer Science and Engineering - University of Bologna - paola.salomoni@unibo.it

1. Problem Addressed

Urban accessibility is a primary factor for social inclusion and the effective exercise of citizenship for everybody, including people with disabilities and elderly people. Moreover, the lack of information about the urban environment and its accessibility represents itself a barrier to those citizens, who are strongly discouraged from step out of well-known paths [1].

This work presents our research related to mPASS (mobile Pervasive Accessibility Social Sensing), a system devoted to collect data about urban accessibility from sensing and crowdsourcing and to provide citizens with personalized and trustworthy maps and paths.

In order to reach this goal mPASS works on heterogeneous information, coming from different sources: sensors, crowdsourcing, official reviews. This approach aims to collect a rich and detailed set of data about urban barriers and facilities, characterized by different levels of trustworthiness. This variety of data can be used to compute maps and paths which are tailored to specific users' needs on the basis of a profile describing both urban and e-accessibility preferences.

2. Background

This work is mainly based on accessibility of maps and navigation services and on user modeling and personalization to support accessibility. Most part of literature on maps accessibility is related to sight impairments and an extensive and well organized review can be found in [2].

Research is not limited to these users, but includes a large set of projects and publications which are devoted to meet special needs related to different disabilities. A review that considers main techniques and methods is available in [3].

Some researches focuses on maps interaction and rendering on mobile devices, such as [4]. In this field, adaptation is used to personalize maps and navigation services, as in [5]. In this context, adaptation is used coupled with user modeling and profiling methods, so as to let user's preferences and special needs drive the customization of interfaces, contents and services. A snapshot of the current state of the art is available in [6].

3. Strategy

We have mainly based our research on these issues: data model, user profile and map adaptation techniques [7].

Our data model is based on accessibility Points of Interest (aPOIs) and reports. An aPOI is an element of the urban outdoor environment (urban design aPOIs) or of an indoor context, such as a restaurant or a public building. We analyzed and grouped in categories more than 200 accessibility barriers and facilities (mainly focusing on the outdoor ones) in the following categories: Gap, Crossing, Obstruction, Parking, Surface, Pathway, Light, Vehicles (public means of transport) and Stations (metro stations, bus stops, etc. [8]). An exhaustive description can be found in [9].

For each category we decided which barrier/facility can be identified by sensing and crowdsourcing and which one would be recognized only by users' reviews (i.e., only by crowdsourcing). For each aPOI, one or more reports are collected.

A report contains information gathered by one of three possible sources: sensors, the crowd and data officially provided by experts:

  • Sensor reports, gathered by smartphones, used as sensors, while users are moving in the urban environment. By exploiting gyroscope, accelerometer and GPS, we can identify simple barriers and facilities (such as steps or ramps).
  • User reports, obtained by crowdsourcing. Smartphones are used to provide data through reviews about urban accessibility. We can collect reviews from users in two ways:
    1. volunteering: users who want to send a review about the place where they are (or a place where they have been);
    2. on demand: users can be asked to produce a review about a place where they are (or nearby).
  • Experts reports, obtained from official reviews done by authorities and organizations (e.g., local administrations, organizations that support people with disabilities, hotels associations, etc.). Experts can assess the actual accessibility of a place, reporting barriers and facilities.

We exploit a XML-based user profile, structured in three interconnected parts:

  1. Generic Profile: it includes some general data (e.g., users personal info, average speed, credibility).
  2. Urban Accessibility Profile: it describes preferences related to each urban accessibility barrier/facility. Users can define each aPOI as [9]:
    • NEUTRAL: The user has neither difficulties nor preferences related to the aPOI type.
    • LIKE: The user prefers aPOIs of this type, when they are available.
    • DISLIKE: The user can face this aPOI type, but with some efforts.
    • AVOID: The user cannot face this aPOI type and an alternative path is necessary.
    On the basis of these preferences, we can compute a route, which comes across the LIKEd aPOIs when feasible, gets round the DISLIKEd aPOIs if possible and totally avoids the AVOIDed aPOIs every time. Users can be equipped with different paths characterized by different lengths and different matches with their preferences. Positive preferences can be associated to barriers and negative preferences can be associated to facilities (e.g., a blind user can set as LIKE some specific barriers, such as stairs and steps, because they can represent a reference point).
  3. e-Accessibility Profile: it is devoted to store preferences and needs in terms of maps rendering.

On the basis of such a profile, it is possible to customize maps to meet users' needs of users. Maps are adapted with the aim of providing users with data related to barriers/facilities which impact on their paths and with accessible information. To reach these goals, we have exploited a set of adaptation techniques, which have been previously used in the fields of accessibility [3] and perceptual cartography [4]. The main adaptation techniques we applied are: Map to text, Exaggeration, Elimination, Typification, Color personalization and Textual detailing. A more detailed description can be found in [10].

4. Major Difficulties

Working on the basis of possibly unreliable reports, our data model can generate both false positives (due to non-existing barriers and facilities detected) and negatives (due to undetected existing barriers and facilities): they can cause difficulties to users in terms of dealing with longer paths or preventing them from reaching their destination. Hence, a trustworthiness assessment has been defined with the aim of evaluating each report on the basis of the source credibility/accuracy, by combining all the reports related to an aPOI, computing an aPOI trustworthiness value. On the basis of this value, an aPOI can be classified as present, absent or uncertain.

In [11] we propose a simple method to compute the trustworthiness of an aPOI and we define a value of sufficient trustworthiness (as a function of the number of reports related to an aPOI). We are conducting simulations to evaluate such a trustworthiness assessment and some initial results are presented in [11].

5. Outcomes

While designing and developing mPASS, we have:

  • adopted the data model described above (including the user profiling);
  • developed and tested some sensing apps (identifying steps and stairs by means of smartphone sensors [1]);
  • designed a specific routing algorithm, based on GraphHopper, with the aim of meeting users' preferences;
  • integrated our data with accessibility reviews made by associations and organizations (e.g., AccessTogether, IngressoLibero) and with OpenStreetMap data;
  • applied a notification module, which sends requests to users on the basis of their location and credibility in order to explicitly ask additional reports (with the aim of improving trustworthiness of aPOIs, trying to solve uncertain aPOIs).

6. Open Research Avenues

This work is still under study and development and several open research issues would be investigated.

A collaboration with the Madeira Interactive Technology Institute is ongoing, with the aim of applying gamification techniques so as to better involve users in data collection.

Thanks to a collaboration with the Emilia Romagna Region, we are integrating our data model with the T-per (the local provider of public means of transport) open data about real time availability of public transportation means, their equipment in terms of accessibility barriers and facilities, their time of arrival and route, etc. This would let us design and provide a multimodal travel planner to citizens with special needs in local urban environments.

References

  1. C. Prandi, P. Salomoni, and S. Mirri (2014) mPASS: Integrating People Sensing and Crowdsourcing to Map Urban Accessibility. Proceedings of the IEEE International Conference on Consumer Communications and Networking Conference (CCNC '14).
  2. L. Zeng and G. Weber (2011) Accessible Maps for the Visually Impaired. Proceedings of 13th IFIP TC13 Conference on Human-Computer Interaction (INTERACT 2011), 54-60. DOI: 10.1007/978-3-642-23768-3_132.
  3. K. Miesenberger (2012) Accessible Maps. W3C Research and Development Working Group Wiki. http://www.w3.org/WAI/RD/wiki/Accessible_Maps.
  4. V. Setlur, C. Kuo, and P. Mikelsons (2010) Towards Designing Better Map Interfaces for the Mobile: experiences from example. Proceedings of the 1st ACM International Conference and Exhibition on Computing for Geospatial Research & Application (COM.Geo '10), Article No. 31. DOI: http://dx.doi.org/10.1145/1823854.1823890.
  5. B. Weninger (2012) Deducing Parameters for Personalizing Maps from Map Interaction Patterns. Proceedings of the 2012 ACM International Conference on Intelligent User Interfaces (IUI '12), 341-344. DOI: http://dx.doi.org/10.1145/2166966.2167044.
  6. W3C/WAI-RDWG (2013) Proceedings of the User Modeling for Accessibility online symposium. http://www.w3.org/WAI/RD/2013/user-modeling/#proceedings.
  7. C. Prandi (2014) Accessibility and Smart Data: the Case Study of mPASS. Proceedings of the 11th Web for All Conference (W4A2014), Article No. 26. Google Student Award. DOI: http://dx.doi.org/10.1145/2596695.2596723.
  8. K. Hara, S. Azenkot, M. Campbell, C. L. Bennett, V. Le, S. Pannella, R. Moore, K. Minckler, R. H. Ng, J. E. Froehlich (2013) Improving public transit accessibility for blind riders by crowdsourcing bus stop landmark locations with Google street view. Proceedings of the 15th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2013), Article No. 16.DOI: http://dx.doi.org/10.1145/2513383.2513448.
  9. S. Mirri, C. Prandi, P. Salomoni, F. Callegati, and A. Campi (2014) On Combining Crowdsourcing, Sensing and Open Data for an Accessible Smart City. Proceedings of the 3rd International Conference on Technologies and Applications for Smart Cities (I-TASC'14).
  10. S. Mirri, C. Prandi, and P. Salomoni (2014) A Context Aware System for Personalized and Accessible Pedestrian Paths. Proceedings of the International Conference on High Performance Computing & Simulation (HPCS 2014), 833-840. DOI: http://dx.doi.org/10.1109/HPCSim.2014.6903776.
  11. C. Prandi, P. Salomoni, and S. Mirri (2014) Trustworthiness Assessment in Mapping Urban Accessibility via Sensing and Crowdsourcing. Proceedings of The First International Conference on IoT in Urban Space (Urb-IoT 2014).