Driver-centric urban logistics optimization: vehicle routing with heterogeneous fixed drivers

Abstract

In the era of flourishing online shopping, efficient delivery operations are essential for companies to enhance customer satisfaction and stay competitive. Previous studies often overlooked the attributes of delivery drivers, particularly their regional knowledge, which significantly impacts the optimization of last-mile delivery operations. This paper introduces the vehicle routing problem with heterogeneous fixed drivers (VRP-HFD), a novel extension of traditional vehicle routing models. It incorporates the unique attributes of drivers, including their regional knowledge and driving qualifications. We propose a mathematical model that captures the relationships among drivers, vehicles, and customers, aiming to minimize both the number of unserved customers and total operational costs. To address the increased complexity in the VRP-HFD model, we develop a hybrid adaptive genetic algorithm (HAGA) that integrates multichromosome coding, adaptive genetic operators, and local search techniques to enhance solution quality. Computational experiments using real-world instances demonstrate the effectiveness of HAGA in generating high-quality solutions compared to the baseline algorithms. Our findings underscore the critical role of driver attributes in optimizing last-mile delivery operations, offering valuable insights for logistics companies committed to improving service efficiency and customer satisfaction.