Mathematical models and optimisation algorithms for dynamic warehouse systems
The overall objective of this doctoral project is to develop mathematical models and optimisation algorithms for dynamic warehouse systems, supporting proactive robot recharging and SKU repositioning operations to improve overall warehouse efficiency across interdependent warehouse processes, including storage location assignment, picking and packing.
More specifically, the project seeks to (i) integrate picking and packing optimisation under realistic operational constraints to establish a benchmark scenario, (ii) design proactive strategies for robot recharging based on stochastic workload and energy usage, (iii) develop online SKU repositioning methods to dynamically adapt warehouse layouts during operations, and (iv) evaluate the robustness and adaptability of these strategies under non-stationary demand patterns, leveraging demand forecasting techniques to inform operational decisions.
The incorporation of forecast-driven algorithms and adaptability to non-stationary demand patterns will enable warehouses to anticipate operational bottlenecks, reduce downtime, and improve throughput. These innovations not only advance the state of the art in operations research but also deliver practical tools and methodologies that enhance reliability, scalability, and responsiveness in e-commerce logistics, ultimately supporting faster delivery times and cost-efficient operations.
The project aims to deliver open-source benchmark instance generators, simulation tools (both exact and heuristic algorithms) for performance evaluation embedded in a simulation environment, and multiple scientific manuscripts detailing the proposed methodologies and their impact on operational efficiency.