From efficiency to experience: Human-centered cyber-physical systems for urban logistics

Cyber-Physical Systems (CPS) tightly integrate sensing, computation, and control to couple the physical and cyber worlds, enabling data-driven modeling and decision-making that enhance real-world operational performance. Driven by the expansion of e-commerce and accelerating urbanization, urban logistics, an ecosystem that generally relies on couriers as core operational agents, has become a critical and rapidly evolving domain for human-centered CPS. The large-scale data generated by the urban logistics system provides an opportunity to apply advanced data-driven modeling and decision-making techniques to effectively improve urban logistics operations. However, a comprehensive human-centered CPS for urban logistics that optimizes operations from multiple perspectives is still lacking.Therefore, this dissertation aims to understand courier behaviors in urban logistics with data-driven methods and then enhance the system performance through intelligent decision-making policies to achieve two key objectives: improving system efficiency and courier experience. To achieve these objectives, this dissertation presents three works. Firstly, to optimize the system efficiency for order delivery, we propose a relay-based last-mile delivery scheme and design REDE, a relay courier scheduling system, which predicts delivery couriers’ mobilities and then optimizes relay courier scheduling for relaying orders from the station to delivery couriers, thereby improving delivery efficiency. Secondly, we extend our work by co-optimizing the system efficiency and the courier experience for order delivery. Specifically, we design WORD, a workload-constrained order assignment system, which measures courier energy consumption of heterogeneous behaviors (i.e., cart pulling, walking, and stair climbing) and then optimizes order assignment with the data-driven measurement results, to achieve fair and efficient delivery order assignment. Finally, we further extend our work from delivery orders to heterogeneous tasks in urban logistics. Specifically, we design HeroCS, a heterogeneous task assignment system that optimizes real-time courier scheduling for delivery, pickup, and customer expansion tasks, considering both efficiency (i.e., task completion rate) and courier experience (i.e., the fairness among couriers). These works are implemented and evaluated with real-world data from one of the largest logistics companies in China. The promising results pave the way towards human-centered CPS for future urban logistics.