Human-Robotic Collaborative Mode Analysis in Human-Robotic Collaborative Order Picking Systems

Human-robotic collaborative order picking system is a flexible and high-efficient warehouse picking system, which integrates the advantages of manual picking system and robotic warehousing system. We are first to propose three collaborative modes, namely Couple, Robot-in-lead, Picker-in-lead, and discuss the resource allocation and zoning strategies in these modes. To investigate the system performance under these modes, we formulate fork-join queueing network models in which an arrive batch orders will be split into robot tasks and picker tasks. We develop an approximation method to estimate the system performance and validate the analytical models by simulation. Numerical experiments are carried out to compare the system throughout time of collaborative modes under different resource allocation and zoning strategies. The results show that, to achieve resource allocation balance, Couple should make the number of robots equal with pickers, Robot-in-lead should make the number of pickers more than robots, and Picker-in-lead only need to make the number of pickers sufficient. Robot-in-lead can benefit from zoning strategy. More zones and pickers will lead to the decrease of the minimal system throughout time, but it does not work in Picker-in-lead. If company does not pursue lowest system throughput time, Picker-in-lead is the best choice. If company is well-capitalized, Robot-in-lead will decrease the system throughout time.