Machine Learning for Automation of Warehouse Activities

Abstract

This paper presents machine learning approaches for automation of activities in warehouses. Integrating machine learning in supply chain management can help automate a number of mundane tasks and allow enterprises to focus on more strategic and impactful business activities. The various machine learning models presented are designed to work with low-cost hardware. The models were studied with different sizes of the input data and the most appropriate ones were selected according to set criteria. Their ability to run on Raspberry Pi single-board computer has been explored and performance characteristics in inference mode have been experimentally established.