Robust Passenger Itinerary Planning Using Transit AVL Data

Automatic vehicle location (AVL) data can be used for a variety of operational and planning purposes at a transit agency. One possible application is for passenger itinerary planning, where the passenger seeks to go from an origin to a destination in some “optimal” way. The AVL data give actual vehicle arrival and departure times, which in turn can give more robust itineraries to passengers. The idea is to use the historical patterns of vehicle arrival and departure times to inform passengers of the actual service. Information on the schedule reliability, likelihood of making a transfer, the range of possible arrival times at the destination, and related service information, could be given to passengers to improve trip planning. A method to determine such robust itineraries is described. This method uses recent techniques for stochastic and timedependent shortest paths to determine a set of possible itineraries for the passenger. To illustrate the method and to illustrate its utility for passengers, a small example using the bus network in Tucson, Arizona is given. Advantages of this method over traditional itinerary planning methods are also outlined.