Optimization of Last-Mile Delivery Using Predictive Order Clustering
Keywords:
Predictive Order Clustering, Last-Mile Delivery Optimization, E-Commerce Logistics, Machine Learning, Vehicle Routing Optimization, Spatial–Temporal ClusteringAbstract
Last-mile delivery represents one of the most complex and cost-intensive components of modern e-commerce logistics. Rapid growth in online shopping has significantly increased delivery volumes, creating challenges related to route efficiency, delivery time windows, and vehicle utilization. This study proposes an intelligent framework for optimizing last-mile delivery using predictive order clustering and data-driven routing strategies. The proposed approach integrates machine learning–based order prediction, spatial–temporal clustering, and vehicle routing optimization to improve delivery efficiency in urban environments. Last-mile delivery represents one of the most complex and cost-intensive components of modern e-commerce logistics. Rapid growth in online shopping has significantly increased delivery volumes, creating challenges related to route efficiency, delivery time windows, and vehicle utilization. This study proposes an intelligent framework for optimizing last-mile delivery using predictive order clustering and data-driven routing strategies. The proposed approach integrates machine learning–based order prediction, spatial–temporal clustering, and vehicle routing optimization to improve delivery efficiency in urban environments. The framework first analyzes historical order data to forecast order arrival patterns and identify spatial delivery trends. Using these predictions, orders are grouped through clustering techniques that consider both geographical proximity and temporal delivery constraints. The clustered orders are then assigned to delivery vehicles through a routing optimization mechanism that minimizes travel distance while satisfying vehicle capacity and delivery time window constraints. Overall, the study highlights the potential of artificial intelligence and predictive analytics in transforming last-mile delivery operations within e-commerce ecosystems. The proposed model provides a scalable and practical solution for modern logistics networks, supporting both operational efficiency and customer satisfaction.
Downloads
References
J. C. Ferreira and M. Esperança, "Enhancing sustainable last-mile delivery: The impact of electric vehicles and AI optimization on urban logistics," World Electric Vehicle Journal, vol. 16, no. 5, p. 242, 2025.
T. Khadra, "A Predictive Model for Improving Last-Mile Delivery: Enhancing Operational Efficiency Through Advanced Analytics The Case of Logistics in Jordan," Princess Sumaya University for Technology (Jordan), 2025.
A. K. Kalusivalingam, A. Sharma, N. Patel, and V. Singh, "Optimizing e-commerce revenue: Leveraging reinforcement learning and neural networks for AI-powered dynamic pricing," International Journal of AI and ML, vol. 3, no. 9, 2022.
Q. Lu, H. Lyu, J. Zheng, Y. Wang, L. Zhang, and C. Zhou, "Research on E-Commerce Long-Tail Product Recommendation Mechanism Based on Large-Scale Language Models," in Proceedings of the 9th International Conference on Electronic Information Technology and Computer Engineering, 2025, pp. 997-1002.
N. Ahmed, M. E. Hossain, Z. Hossain, M. F. Kabir, and I. S. Hossain, "Assessing the Potential and Ethical Implications of Agentic AI in Surveillance Technology," Formosa Journal of Multidisciplinary Research, vol. 4, no. 4, pp. 1841-1858, 2025.
I. Johansen, "AI-Augmented API Management: Enhancing Security, Governance, and Performance in Enterprise Ecosystems," 2023.
L. Bhavani Karanam, "AI-Driven Optimization of Last-Mile Delivery: Reducing Costs and Enhancing Efficiency in Urban Logistics," 2025.
K. L. Sainani, "Explanatory versus predictive modeling," PM&R, vol. 6, no. 9, pp. 841-844, 2014.
Y.-H. Chen, Y.-Z. Li, H. Jiang, and Z. Huang, "Research on household energy demand patterns, data acquisition and influencing factors: A review," Sustainable Cities and Society, vol. 99, p. 104916, 2023.
S. M. J. Mirzapour Al-e-Hashem, T.-H. Hejazi, G. Haghverdizadeh, and M. Shidpour, "Optimizing last-mile delivery services: a robust truck-drone cooperation model and hybrid metaheuristic algorithm," Annals of Operations Research, pp. 1-31, 2024.
Ö. Aslan Yıldız, İ. Sarıçiçek, and A. Yazıcı, "A reinforcement learning-based solution for the capacitated electric vehicle routing problem from the last-mile delivery perspective," Applied Sciences, vol. 15, no. 3, p. 1068, 2025.
