Shortcut suggestion based on collaborative user feedback for suitable wheelchair route planning

Traditionally, available route planners suggest paths in terms of streets, although, sidewalks of a same street may present different accessibility conditions for wheelchair users. To address this problem, we describe in this paper a sidewalk-based model for wheelchair route planning. The model is a graph in which vertices are corners that makeup city blocks and edges are sidewalks or pedestrian crossings. Namely, we automatically build a coarse graph from geographic information system maps (endorsed by discrete features, such as points, lines, polygons). Then, based on user feedback, the graph model is refined so as to indicate the wheelchair-accessible sidewalks and crosswalks. To evaluate our model we considered a neighborhood in a Brazilian city. However, the user feedback have shown severe issues to the city connectivity, mainly due to the lack of curb ramps. We show that such issues can be reduced by adding k shortcut edges, suggested by a greedy algorithm, in a refined graph, so as to minimize the weighted average shortest path distance over all pairs of vertices.