Exploring the Applicability of Machine Learning in Assessing Last Mile Delivery Route Properties
Date:
Presented research on machine learning applications for analyzing and predicting last-mile delivery route characteristics at the INFORMS Annual Meeting in Seattle.
Presentation Overview
This talk discusses a comprehensive approach to using machine learning for understanding and predicting properties of last-mile delivery routes. The research analyzes real-world logistics data to identify factors affecting route deviations and delivery performance.
Key Points
- Analysis of 20,000+ historical delivery routes
- Machine learning models for route deviation prediction
- 19% improvement in prediction accuracy over baseline methods
- Practical implications for logistics operations and planning
Impact
The findings help logistics companies better understand and manage last-mile delivery challenges, enabling more accurate route planning and improved operational efficiency.