Stochastic Models for Order Picking Systems Defended on Thursday, 8 September 2016
Order picking, the process of retrieving customer orders from their storage locations, is the most critical operation in a warehouse. Any under performance in order picking can lead to unsatisfactory service and high operational cost for the warehouse, and consequently for the whole supply chain in which the company operates. This thesis develops new stochastic models for the performance evaluation of two state-of-the art order picking systems: zone picking and polling-based milkrun picking. These models adequately describe and predict the consequences of variability on the performance of these warehousing systems.
The first part of the thesis zone picking systems are studied, one of the most popular conveyor-based picker-to-parts order picking methods used in practice. We model the various elements of the system including conveyor merges as a network of queues with multiple order classes, with capacity constraints on subnetworks, and with the dynamic block-and-recirculate protocol. The resulting model is most suitable to support rapid and optimal design of complex zone picking systems. In the second part of the thesis, milkrun picking systems are investigated. In this system an order picker picks multiple orders that arrive in real-time and integrates them in the current picking cycle. This subsequently changes dynamically the stops on the order picker’s picking route. Using polling models, we study order throughput times for various picking policies, and the effect of product allocation. The results of the model show that when the order arrival rate is high milkrun order picking significantly improves system performance compared to conventional batch picking. In addition, the best product allocation improves the order throughput time considerably.
Warehousing, Queueing Models, Stochastic models, Optimization, Order picking, Zone picking, Milkrun picking