MODEMS: A Tailored Solution for Promoting On-Campus Passenger Mobility

MODEMS ( M unich O n- D emand E lectric M obility S ervice) showcases a Proof-of-Concept (PoC) for enhancing on-campus mobility through the integration of a fleet of Electric Vehicles (EVs), an operational route planning algorithm, and an interactive online portal. The framework enables users to submit transportation requests between designated service stations across the university campus and subsequently notifies them of the request acceptance or rejection based on system availability. Each request comprises a pickup and a delivery location, a passenger count, and a desired service time window at either pickup or delivery. Fleet agents are primarily characterized by their maximum passenger capacity and battery State-of-Charge (SoC) constraints. The core objective is to fulfill all user requests within their respective time windows, while minimizing the cumulative fleet travel time. The problem represents an advanced variant of the multi-agent Capacitated Vehicle Routing Problem with Pickup, Delivery and Time Windows (CVRPPDTW) constraints, for which we present three Mixed-Integer Linear Programming (MILP) formulations. Due to the unpredictability of MILP solution times, we also implement an Adaptive Large Neighborhood Search (ALNS) algorithm as a supplementary strategy to guarantee the generation of routing plans within reasonable time limits. All strategies are evaluated in curated experiments to validate their operational efficacy in realistic scenarios. Finally, we discuss the achieved results, followed by exploring potential future research directions.