A Systematic Literature Review of Last-Mile Delivery: Analyzing Current Situations and Identifying Research Gaps for Future Perspectives

Abstract

Last-mile delivery (LMD) has become a critical challenge in logistics, driven by the rapid rise of e-commerce and growing consumer expectations for speed, flexibility, and sustainability. Despite significant advances in optimization models, simulation techniques, and machine learning applications, existing reviews remain fragmented and offer limited comparison of methodological strengths. This study addresses this gap through a systematic literature review of LMD research published between 2019 and 2025, following the PRISMA framework. Out of 162 screened publications, 33 were analyzed using bibliometric mapping in R and methodological clustering in Python. The analysis identifies four methodological streams: optimization and mathematical modeling, simulation and heuristics, multi-criteria and learning models, and literature-based analytical studies. Beyond classification, the study highlights critical gaps, including the limited integration of behavioral factors, financial trade-offs, and policy implications. By mapping methodological trends and proposing targeted research directions, this work provides a foundation for advancing more resilient, adaptive, and sustainable last-mile delivery systems.