Supermarket Shelf Monitoring Using ROS based Robot

This paper presents a robotic system that automates the process of keeping a check of product quantity in the store shelves of the supermarkets and thereby taking relevant refilling action in case of product shortage. Due to the high manpower costs associated with the current traditional approach to shelf monitoring, these activities are not carried out as frequently as they should be, resulting in increased customer churn and revenue loss for supermarket owners. For this, a ROS based autonomous robot is developed to make it easier to track and spot out-of-stock goods. For a robot to be autonomous in every aspect, it should be efficient, flexible, and have a reliable navigation system, which can be achieved by using Robot Operating System (ROS) framework. This work produces virtual robots, maps, and sensors on simulations in the GAZEBO application, a virtual environment that helps in simulating the dynamics and physics of the robotic systems. For navigation and mapping, the Gmapping particle cloud algorithm is used. In addition to this, OpenCV, a python package for image analysis and object detection, and RVIZ tool for visualization purposes is used for the proposed system.