Robotic cellular warehousing systems : introduction, operations, and management

Abstract

With the growing popularity of e-commerce, logistics operations have been challenged by the higher demand for and increased complexity of order picking in warehouses. While goods-to-person picking systems, such as robotic mobile fulfillment systems, are becoming prevalent, these systems still have shortcomings, including the low-effective robot transportation and unstable performance of manual picking due to fatigue and human errors. To address these issues, this thesis introduces a robot-to-goods order picking system named RubikCell. Applying cellular warehousing principles to RubikCell, which forms the Robotic Cellular Warehousing System, makes e-commerce warehousing more flexible, scalable, and reconfigurable. The entire thesis focuses on this innovative system and explores it at the operational, tactical, and strategic levels.

At the operational level, a multi-robot parallel picking scheme is proposed and formulated, which involves three stages of planning: (i) the construction of order routes, (ii) order-to-robot allocation, and (iii) multi-robot conflict-free path planning. Particularly, a time-space-state network (TSSN) model is developed that takes into account three-dimensional information to identify feasible and efficient paths for robotic order picking. Numerical experiments showcase the effectiveness of the proposed approach and determine optimal system configurations, including the length-to-width ratio of a warehousing cell, the number of robots within a warehousing cell, and the cell size.

At the tactical level, different operating policies are examined. Two picking strategies (pick-while-sort and pick-then-sort) and three robot-to-workstation assignment rules (random, closest, and dedicated) are considered. Single-class and multi-class closed queuing network (CQN) models are established to estimate warehouse throughput capacity under different policies. Numerical simulations and experiments are conducted to investigate the impact of the numbers of robots and workstations, robot capacity, order size, and sorting efficiency on warehouse throughput capacity.

At the strategic level, the warehousing mode (dedicated warehousing and shared warehousing) and the service level are analyzed using game-theoretical models. The findings can determine the warehousing mode choice of the third-party warehousing service provider, the e-tailer’s preferences for the warehousing mode, all-win scenarios, and supply chain performance. These insights contribute to enhancing the efficiency and profitability of the warehousing industry and facilitating collaborative partnerships within the supply chain.

Overall, this thesis aspires to alleviate the bottleneck in e-commerce logistics by addressing challenges in order picking through the introduction of the Robotic Cellular Warehousing System. It also provides managerial implications at operational, tactical, and strategic levels.