Personalized Travel Time Prediction Using a Small Number of Probe Vehicles

Predicting the travel time of a path is an important task in route planning and navigation applications. As more GPS probe data has been collected to monitor urban traffic, GPS trajectories of the probe vehicles have been frequently used to predict path travel time. However, most trajectory-based methods rely on deploying GPS devices and collect real-time data on a large taxi fleet, which can be expensive and unreliable in smaller cities. This work deals with the problem of predicting path travel time when only a small number of cars are available. We propose an algorithm that learns local congestion patterns of a compact set of frequently shared paths from historical data. Given a travel time prediction query, we identify the current congestion patterns around the query path from recent trajectories, then infer its travel time in the near future. Driver identities are also used in predicting personalized travel time. Experimental results using 10--25 taxis in urban areas of Shenzhen, China, show that personal prediction has on average 3.4mins of error on trips of duration 10--75mins. This result improves the baseline approach of using purely historical trajectories by 16.8% on average, over four regions with various degrees of path regularity. It also outperforms a state-of-the-art travel time prediction method that uses both historical trajectories and real-time trajectories.