Dynamic location for charging operations of shared free-floating e-scooters

Abstract

Shared electric scooters (e-scooters) have recently become a popular mode of micromobility solution and this rapid growth causes significant operational challenges. One of the challenges micromobility companies face is collecting idle low-charge fleets from different corners of the city. The collector is usually a truck that needs to make a tour in town to collect these e-scooters. One solution can be for the operator to dynamically define special zones where e-scooters with low-charge are collected. In this paper, we propose a two-layered approach to tackle the problem. We propose a dynamic programming approach to investigate the required number and locations of designated low-charge drop-off points for the first layer of the problem. A simulated origin-destination data set of 30 e-scooters on TU Delft campus area is used for a time horizon of 8 working days and a time discretization level of 15 minutes. The results suggested consolidating low-charge e-scooters in three low-charge drop-off locations considering the state of charge (SOC) and distance of e-scooters from the low-charge drop-off zones. Then, we discuss the potential implications of our findings on recharging operations and spatial efficiency that are defined dynamically.