Edge-Based Infrared-Ultrasonic Anti-Collision Radar System for Robotic Navigation: *Applications of Cost-effective Bisensory System for Obstacle Detection, Tracking, and Avoidance

Abstract

Significant progress has been made in the development of an edge computing system for the fusion, representation, and visualization of dual-sensor data for obstacle avoidance control of mobile robots. This involves the use of an error-filtering covariance and averaging algorithm to logically fuse distance measurements from a pair of infrared and ultrasonic sensors, which was instrumented into a robot and used to generate radar visuals to track the proximity of ambient obstacles within 180 degrees spanning ahead of the robot. Hands-on experiments were performed to evaluate the performance and applicability of the system in real-time. The results show that the developed system is viable and robust. In line with the emerging field of edge computing, this work is an efficient, portable, and cost-effective approach to developing mobile robotic systems.