Detecting Deviations from Intended Routes Using Vehicular GPS Tracks

摘要

This article proposes a method to find intersections at which cars tend to deviate from the optimal route based on global positioning system (GPS) tracking data under the assumption that such deviations indicate that car navigation systems (CNSs) and road signage are not readily available. If the intended route is known, deviations can be enumerated by comparing the intended route with the vehicle's actual route as observed by a GPS; however, the intended route is unknown and can differ from the route suggested by a CNS. To identify intersections with high deviation rates without knowing intended routes, we exhaustively sampled subsequences from each vehicular GPS track, and detected deviations from the optimal route for the subsequences. Although the detected deviations are not always caused by driver confusion, accumulating such erroneous detection results would yield a meaningful difference in the number of accumulated deviations at each intersection. We applied the proposed method to 3,843 GPS tracks collected from visitor drivers in the city of Kyoto. Thresholding the estimated deviation rate yielded 39 intersections from 14,543 candidates. The results show a certain level of correlation between obtained deviations and rerouting locations from actual CNS data. We also found several intersections where faulty route suggestions are provided by CNSs.