LAST-MILE DELIVERY WITH ELECTRIC VANS: TIME-WINDOW ROUTING UNDER BATTERY DEGRADATION UNCERTAINTY

Abstract

The last-mile delivery (LMD) is an important stage of logistics, it has a very strong impact on quality of service, costs and environment. E-commerce has turned electric vans into a promising alternative to LMD challenges, though concerns such as battery depreciation variability further complicates scheduling and routing. This paper studies the impact of battery degradation uncertainty on time-window routing in last-mile distribution problems with electric vans by aiming to design an optimal strategy that can trade-off between time-efficiency and energy constraints. The work utilizes a hybrid methodology of simulation and optimization to investigate time-window routing for electric vans with uncertainty in battery degradation. A mathematical formulation is put forward in which the stochastic nature of battery consumption rates is combined with time-window delivery requirements. The evaluation is performed on a synthetic dataset that models a typical urban area with 100 delivery location, 10 delivery vans and time window in the middle of interval [30,120] minutes for each newspaper. The model demonstrates that accounting for battery degradation uncertainty in routing reduces delivery lateness by 15% and extends the lifetime of an EV's battery by 10%, outperforming deterministic models. Logistics and fleet managers can leverage the model to minimize last-mile delivery routes and to mitigate environmental effects, as well as to mitigate their operational waste. This is the first empirical study on last-mile delivery in a battery degradation uncertainty context, which contributes both theoretically and practically to logistics and operations management.

Keywords: Last-mile delivery, electric vans, battery degradation, time-window routing, stochastic optimization.