Multi-objective optimization of two-echelon delivery with autonomous delivery vehicles and electric two-wheelers

Couriers often make multiple round trips between warehouses and customer locations within short time frames in traditional last-mile delivery. A two-echelon delivery system that integrates autonomous delivery vehicles (ADVs) and electric two-wheelers (E2Ws) holds significant promise for minimizing detours and reducing empty mileage, thereby enhancing delivery efficiency and lowering labor costs. Consequently, we propose a two-echelon delivery problem for the routing and scheduling of ADVs and E2Ws, considering the synchronization reliability for two-echelon vehicles and the system cost, greenhouse gas emissions, and delivery risk. A hybrid algorithm integrating an adaptive large neighborhood search heuristic and a multi-objective genetic algorithm is designed to address the problem. Experiments in Shanghai reveal that our proposed model could reduce system costs (27.11%) and emissions (30.62%) when compared to traditional three-wheeler delivery. The multi-objective setting exhibits a substantial decline in delivery risk by above 20%. Finally, the stochastic simulation approach validates the superior performance of the synchronization reliability constraint.