A deep learning approach for the selection of an order picking system

This paper develops a novel strategic decision support framework for the design of an order picking system, which can be used whenever different systems and/or control mechanisms need to be compared for a given customer order structure. Warehousing companies frequently struggle in selecting the most suitable design for their order picking system. Traditionally, a comparison of different order picking systems is based on time-consuming simulation runs. In addition, the only source of consultancy is most often carried out by the order picking system manufacturers themselves. Our framework, using recent advancements in deep neural networks, provides an efficient methodology for selecting not only the best order picking system for a given order structure but also the most suited design parameters. This enables warehouse companies to compare objectively an extensive number of systems and allows them to identify the most promising order picking systems. We demonstrate our framework for a comprehensive comparison of three different fixed-path order picking systems to find one best suited for a provided order structure. (C) 2022 Elsevier B.V. All rights reserved.

Automated summary

This paper presents a novel strategic decision support framework for the design of an order picking system, which allows for the comparison of different systems and control mechanisms in a given customer order structure. By utilizing recent advancements in deep neural networks, the framework provides an efficient methodology for selecting the best order picking system and design parameters. It enables warehouse companies to objectively compare a large number of systems and identify the most promising order picking systems.