Design of a Lightweight Obstacle Detection System for Mobile Robot Platforms with a LiDAR Camera

The treatment of patients in isolation is one of many new challenges that doctors and medical staff have to deal with daily. Thus, recent developments in robotics for the healthcare industry have led to many researchers building autonomous robots to assist in performing menial and labor-intensive tasks, such as delivering items to and from hospital rooms. To develop an indoor navigation system, these robots need a robust obstacle-detection method to guide the robot safely. In this paper, we present a novel and lightweight method to detect obstacles with a Light Detection and Ranging (LiDAR) depth camera for any mobile robot platform that uses the Robot Operating System (ROS) framework. The camera generates a depth image projection of the closest objects in the form of point cloud data. This data is converted into planar data to reduce the computation demand and simplify the incoming depth image data for collision avoidance in a navigation system. The results proved to be highly accurate with the LiDAR camera scanning in three different scenarios.