Computing and Evaluating Multimodal Journeys

We study the problem of finding multimodal journeys in transportation networks, including unrestricted walking, driving, cycling, and schedule-based public transportation. A natural solution to this problem is to use multicriteria search, but it tends to be slow and to produce too many journeys, several of which are of little value. We propose algorithms to compute a full Pareto set and then score the solutions in a postprocessing step using techniques from fuzzy logic, quickly identifying the most significant journeys. We also propose several (still multicriteria) heuristics to find similar journeys, but much faster. Our experiments show that this approach enables the computation of high-quality multimodal journeys on large metropolitan areas, and is fast enough for practical applications.