Efficient Algorithms for Fully Multimodal Journey Planning

We study the journey planning problem for fully multimodal networks consisting of public transit and an arbitrary number of non-schedule-based transfer modes (e.g., walking, e-scooter, bicycle). Obtaining reasonable results in this setting requires multicriteria optimization, making the problem highly complex. Previous approaches were either limited to a single transfer mode or suffered from prohibitively slow running times. We establish a fully multimodal journey planning model that excludes undesirable solutions and can be solved efficiently. We extend existing efficient bimodal algorithms to our model and propose a new algorithm, HydRA, which enables even faster queries. On metropolitan and mid-sized country networks with walking and e-scooter as transfer modes, HydRA achieves query times of around 30 ms, which is fast enough for interactive applications.