A Low-cost Localization System for Warehouse Inventory Management

With the increasing demand of e-commerce customers, fast picking and delivery services in a warehouse are required to meet their requirements. One way to facilitate the shipment of orders without incurring additional labor costs is to deploy warehouse robots for moving packages. However, tracking the movement of warehouse robots accurately is not a trivial task. In the past, attempts were made by using line following robots. This method only worked in the absence of intersecting lines. In order to avoid this problem, our approach to track the robot is based on indoor localization using Wi-Fi. However, the deployment of this system is considered difficult compared to the Global Navigation Satellite System (GNSS) due to the non-line of sight and multi-path issues. In this paper, an indoor localization system is proposed to localize and track warehouse robots for inventory management. To reduce deployment cost and increase the adaptability, existing Wi-Fi infrastructure is used for localization. The warehouse robot prototype is designed using a low-cost microcontroller, Inertial Measurement Unit (IMU) sensors and wheel encoders. Experiments show that by applying Kalman Filter for sensor fusion, the system is able to track the movement of a robot accurately using low-cost sensors.