|M.Sc Student||Cohen Achituv|
|Subject||Building a Weighted Graph based on OpenStreetMap Data for|
Routing Algorithms for Blind Pedestrians
|Department||Department of Civil and Environmental Engineering||Supervisors||Dr. Sagi Dalyot|
|Dr. Shiri Azenkot|
|Full Thesis text|
Wayfinding refers to actions people take to successfully complete a navigation process. Since wayfinding is mainly constructed through visual channels, blind pedestrians experience numerous unique challenges. For example, they are not able to see important landmarks, which are essential to maintain a reliable wayfinding process. Moreover, physical objects, such as on-road obstacles and shared lanes with bikes and cars, limit their mobility and pose danger.
This research aims to compute a weighted graph that is based on geospatial environmental data for route calculation designed specifically for blind pedestrians. The weighted graph will enable to compute optimized routes that, instead of being simply shortest in length, would fit blind pedestrians’ needs in terms of accessibility and safety, enabling them to orient and navigate more easily. The road network used is based solely on OpenStreetMap (OSM) data, which is mostly volunteered mapping data collected by the public (i.e., crowdsourced participatory mapping). Since OSM is an open-source mapping platform, it enables insertion and editing of data that are valuable for blind pedestrian navigation and wayfinding, suggesting flexibility in terms of data maintenance and analysis. Still, although volunteered-based map services show an increasing location accuracy, completeness and update-rate of their mapping infrastructure, pre-processing of the road network data is still required to handle the development and implementation of the route calculation algorithm.
This research presents the knowledge gained in interviews, observations and experiments made with blind people and Orientation and Mobility specialists, that lead to the definition of the criteria used that expresses the needs and preferences of blind pedestrians while navigating, namely: length, complexity and type of roads, together with the existence of physical landmarks. The developed algorithms and software implementation were tested and analysed in two cities with different users. Analyses proved that for most cases the proposed criteria used for route calculation expresses correctly the needs and preferences of blind pedestrians while navigating, computing with accessible and safe routes that were easier to orient when compared to the shortest routes. OSM data has the potential to be an effective and valuable mapping database for this task, at least in urban cities in developed countries. This research holds great social contribution to the community of the blind, which can lead to a significant change in mobility and accessibility in the urban areas, contributing to the equality and independence of this community.