Data-driven decision support system for managing item allocation in an ASRS: A framework development and a case study

Abstract When dealing with Automated Storage and Retrieval Systems (ASRS), the allocation of items to the most convenient storage location depends on the vast amount of data produced internally (e.g., Enterprise Resource Planning, Manufacturing Enterprise Systems) and externally (e.g. Supply Chain Management). Moreover, a proper item allocation in the warehouse has a strong influence on the warehouse saturation levels and picking times. In this perspective, the present work proposes the application of data-driven algorithms for managing items in an Automated Storage and Retrieval System (ASRS) in order to reduce the picking times and storage space. Specifically, a four-layer framework is adopted for collecting data produced by different information sources and analyzing them through a data-driven approach. The analytics layer is performed by combining the Association Rule Mining (ARM) technique, to investigate the network of influences among data collected, and a simulation approach for assessing the feasibility of the proposed implementation. The Association Rule Mining allows company managers to identify the components that should be located on the same tray in the ASRS, defining the couples of items frequently picked together in order to reduce the total picking time. The proposed approach is applied to the case study of a shoe manufacturing company to explain the research approach and show how the implementation of the data-driven methodology can provide valuable support in defining item allocation and picking rules. The proposed Association Rule Mining method is new in this context and it has shown a positive impact in comparison to traditional solutions of warehouse management, providing a complete overview of the items’ interactions and identifying communities of items that define local and global patterns and locate influential entities.