[Back]

C. Gaisbauer, S. Winter:
"Strategies for Decentralized Ad-Hoc Trip Planning";
Journal of Spatial Science, 54 (2009), 2; 1 - 14.

This paper considers a set of vehicles, called hosts, forming a transportation network, and a set of mobile transportation clients such as pedestrians. The paper focuses on the problem of clients choosing rides with hosts in an ad-hoc manner, i.e., without calling a central service. For this purpose, clients and nearby hosts are enabled to negotiate via short distance radio, making it likely that clients will not find a complete ride to their destination instantly. Instead, they take rides to intermediate locations, in expectation of finding connecting rides from there on. This paper studies the clients´ dilemma to assess intermediate locations. It proposes a heuristic that outperforms the strategy of strictly following a predefined route (for example, the geometric shortest one). Furthermore, the proposed heuristic is expanded to incorporate a-priori data about the spatially inhomogeneous - yet hosts as it is typical in urban traffic. It is shown that the average travel times of clients can be further reduced through the prediction of host availability at intermediate locations.

LBS, shared ride trip planning, geosensors, heuristics, location ranking, traffic simulation.