Artificial Intelligence and optimization techniques applied to the full truckload pickup and delivery problem

• Grant number: 25/14037-0
• Support Opportunities: Scholarships in Brazil - Master
• Start date: September 01, 2025
• End date: July 31, 2026
• Field of knowledge: Engineering - Production Engineering - Operational Research
• Principal Investigator: Antônio Augusto Chaves
• Grantee: Paulo Roberto Costa Pedro
• Host Institution: Instituto de Ciência e Tecnologia (ICT). Universidade Federal de São Paulo (UNIFESP). Campus São José dos Campos. São José dos Campos , SP, Brazil
• Associated research grant: 24/08848-3 - Full truckload pickup and delivery problem: optimization methods and practical constraints of the supply chain middle mile, AP.R

Abstract

This project aims to investigate and develop efficient solutions to the full truckload pickup and delivery problem in the middle-mile transport of e-commerce companies. The research will address logistical challenges such as heterogeneous and sustainable fleets, charging stations, time windows, and the use of outsourced fleets. Optimizing this stage is essential to reduce costs, improve operational efficiency, and minimize environmental impacts. To this end, exact and heuristic optimization techniques will be applied, with an emphasis on the use of metaheuristics based on random keys and evolutionary algorithms. In addition, the project will explore Artificial Intelligence techniques to optimize these metaheuristics, investigating how LLMs (Large Language Models) can configure the parameters of metaheuristics and how Machine Learning and Deep Learning techniques can approximate the calculation of the objective function, reducing computational cost. It is expected that the methods developed will contribute to making transportation between distribution centers more agile, economical, and sustainable.