Visual marker-guided mobile robot solution for automated item picking in a warehouse

Abstract

To keep up with the boom in e-commerce, fulfillment companies need to leverage robotics and automation to speed up delivery of orders to clients. In this paper, we present a mobile robot with suction grippers designed for item picking in an e-commerce fulfillment warehouse. The robot uses a novel visual marker that we have developed called VITag, to navigate towards the location of the items to be picked, identify them, and generate plans how to approach them. Through the marker, the robot finds the optimum path to the items and the most efficient way to pick them. To improve picking efficiency, the robot derives a picking list (order of how to pick the items) from the order list (items to pick) and its knowledge of the shelf content. The SMACH library for ROS is also used to enable the robot to do high level planning more quickly. Simulation and initial tests show the feasibility of the system.