Optimizing Retrieving Performance of an Automated Warehouse for Unconventional Stock Keeping Units

In recent years, the diffusion of automated warehouses in different industrial sectors has fostered the design of more complex automated storages and handling solutions. These circumstances, from a technological point of view, have led to the development of automated warehouses that are very different from the classic pallet Automated Storage and Retrieval Systems (AS/RS), both in terms of design and operating logic. A context in which these solutions have spread is the steel sector. Warehouses with innovative layouts and operating logics have been designed to move metal bundles of different sizes, weights and quality levels, instead of standard, interchangeable stock keeping units. Moreover, picking is often not allowed in these warehouses, due to the configuration of the loading units. In this work we propose a meta-heuristic algorithm based on the Simulated Annealing (SA) procedure, which aims to optimize performance during the retrieving phase of an automated warehouse for metal bundles. The algorithm translates the customers’ requests, expressed in terms of item code, quality and weight into a list of jobs. The goal is to optimize the retrieving performance, measured in missions per hour, minimizing the deviations in quality and weight between customer request and the material retrieved. For the validation, a simulation model of an existing warehouse has been created and the performance of the algorithm tested on the simulation model has been compared with the current performance of the warehouse.