Operational Micromobility Rebalancing through Crowdsourced Trucks

Abstract

This paper proposes a city-scale framework for station-based micromobility rebalancing that combines censored-demand reconstruction with hybrid truck coordination. While prior studies have examined operator fleets, uncertain demand, and some crowdsourced relocation, city-scale frameworks that jointly combine censored-demand reconstruction, real delivery-truck trajectories, stochastic participation, and rolling MPC remain limited. To address this gap, the paper introduces Shortage-Aware Tobit Attention (SATA) to recover latent demand when low bike inventory suppresses observed departures. The reconstructed demand is converted into shortage and surplus candidates and optimized through a hybrid service-MPC that coordinates deterministic operator trucks with stochastic crowdsourced delivery vehicles. Relative to the operator-only baseline, the full stack achieves its strongest service-rate gains in Chicago first-response stress scenarios and Lyon crowd-advantage settings, while profit gains are concentrated in Lyon.