Next-day demand based last-mile delivery route optimization with collection of reusable bags

The rapid growth of e-commerce has led to increased logistics volumes and widespread use of disposable parcel boxes, raising significant environmental concerns. In response, reusable bags have emerged as a sustainable packaging alternative. However, their benefits can be fully realized only through efficient collection operations. This study proposes a mathematical model for last-mile delivery route optimization that integrates both delivery and reusable bag collection. To address the model’s computational complexity, we develop a Cross-Entropy-based heuristic algorithm capable of generating high-quality solutions within short computation times. Our computational experiments provide key insights. An appropriately set unit penalty cost ensures sufficient bag collection, reduces reliance on disposable boxes, and helps maintain manageable courier workloads. Furthermore, the minimum collection rate significantly impacts both logistics costs and labor requirements, necessitating careful calibration to avoid inefficiencies. A real-world case study based on data from Dongjak-gu, Seoul, demonstrates the scalability and effectiveness of the proposed model and algorithm. The overall framework is highly generalizable and can be readily adapted to various urban logistics contexts, offering practical guidance for designing sustainable, cost-efficient, and courier-friendly last-mile delivery systems.