Improving last-mile delivery operations of electric vehicles using on-demand portable chargers

We investigate the utilization of portable chargers (PCs) for recharging delivery electric vehicles (EVs) en route. While EVs deliver customer orders within predefined time windows, PC delivery vans (PCDVs) are dispatched to supply PCs to EVs at designated customer locations during their visits. We refer to the arising problem as the Electric Vehicle Routing Problem with Time Windows and On-Demand Portable Chargers (EVRPTW-PC). The objective is to minimize overall operational costs by optimizing the fleet size. This study builds upon the Electric Vehicle Routing Problem with Time Windows and Mobile Charging Stations (EVRPTW-MCS) recently proposed in the literature, which primarily focused on recharging EVs at specific locations via Mobile Charging Stations (MCSs). While an MCS should stay parked while recharging an EV, PCDV can deliver the PC and continue its tour to serve other EVs, which improves its overall utilization. In addition, parking restrictions may limit the recharging service of MCSs. Initially, we present the EVRPTW-PC mathematical model of. Then, we propose a matheuristic approach that combines Variable Neighborhood Search alongside an exact method. Finally, we perform numerical experiments to compare the solutions obtained by using PCs and MCSs, and elaborate on the potential benefits of employing PCs.