Two-echelon Electric Vehicle Routing Problem in Parcel Delivery: A Literature Review

TL;DR

This literature review analyzes the two-echelon electric vehicle routing problem in parcel delivery, highlighting current research, challenges, and future directions for sustainable urban logistics involving EVs and multi-echelon systems.

Contribution

It systematically categorizes existing studies on 2E-EVRP, identifies research gaps, and proposes future research directions including advanced modeling and technological integration.

Findings

Identified key research gaps such as tardiness and environmental trade-offs.

Classified 2E-EVRP literature based on variants, objectives, and methods.

Suggested future directions including autonomous vehicles and real-time decision-making.

Abstract

Multi-echelon parcel delivery systems using electric vehicles (EVs) are crucial for managing urban logistics complexity and promoting sustainability. In multi-echelon systems, particularly within two-stage systems, larger vehicles transport parcels from a central depot to satellite hubs, where smaller EVs pick up the parcels and carry out last-mile deliveries. This system could increase efficiency, reduce emissions, and improve service reliability. The two-echelon electric vehicle routing problem (2E-EVRP), an extension of the traditional two-echelon vehicle routing problem (2E-VRP), addresses EV-specific challenges such as battery constraints and recharging stations to tackle environmental impacts, urban congestion, and e-commerce demands. While effectively reducing costs, energy use, and emissions, the 2E-EVRP faces modeling challenges due to multi-echelon structures, EV limitations, and recharging station selection. This paper systematically reviews 2E-EVRP literature, analyzing key studies. It proposes a classification scheme to categorize the papers based on the problem variants, objectives, constraints, and solution methods. It identifies gaps such as delivery tardiness, environmental trade-offs, multi-objective optimization, multiple depots, split deliveries, and time-dependent travel conditions. Future research directions include aligning models with urban policies, integrating parcel lockers, enabling same-day delivery, and incorporating advanced technologies like autonomous vehicles. Methodological advancements suggest using machine learning, reinforcement learning, and simulation-based approaches to enhance dynamic routing and real-time decision-making. These directions aim to expand the 2E-EVRP applicability, addressing theoretical and practical challenges in sustainable urban logistics for future works.