Machine Learning Applied to Last Mile Operations: Applying Machine Learning Models for Stops Classification in Urban Logistics

This study analyzes urban freight GPS data to achieve more efficient public policies and better last-mile operations planning. The main objective is to obtain the time stops of each urban freight vehicle and classify the stop as traffic, delivery, or rest. After classifying each stop using data mining algorithms and applying machine learning models such as K-Means and HDBSCAN, the study demonstrates that it is possible to apply machine learning models capable of grouping GPS stops by similar features. The clusters obtained are relevant inputs to future research and potential applications, such as in public policy decision making or routing optimization allowing private and public entities to optimize urban logistics and last-mile operations.