INTEGRATING CLUSTERING AND ARTIFICIAL INTELLIGENCE FOR IMPROVED EFFICIENCY IN LAST-MILE LOGISTICS

Abstract

The article discusses the problem of optimizing decentralized last-mile delivery in the context of growing e-commerce. An approach based on the use of local microhubs and the application of delivery object clustering methods to reduce the load on central warehouses and shorten transportation time is proposed. A mathematical model has been developed that combines clustering with the task of routing vehicles with time windows. The model takes into account the criteria of minimizing costs, reducing delivery time, and reducing environmental impact. For technical implementation, cartographic and weather APIs are used, as well as traffic forecasting systems, which provide adaptive management in real time. The results of the study confirm that the integration of clustering with artificial intelligence algorithms increases the efficiency of decentralized logistics systems, reduces operating costs, and contributes to the creation of environmentally sustainable and customer-oriented delivery services.