Implementation of Robot Navigation Based on Artificial Potential Field for a Four Omni-directional Soccer Robot

Abstract

The Indonesian Robot Contest (KRI) is an event by the National Achievement Centre, emphasizing design and engineering. The Indonesian Wheeled Robot Soccer Contest is among its divisions, where robots compete to score goals while adhering to rules penalizing collisions. To tackle this, an additional algorithm utilizing the potential field method is crucial for penalty avoidance. It is adept at maneuvering around static and dynamic obstacles, fitting the soccer robot's needs. The potential field algorithm was successfully implemented on OMR kinematics with a coordinate error of less than 3%. Researchers tested static and dynamic obstacles to test the repulsive potential field. For static obstacles, the robot reached the target point with an error of less than 1%, successfully avoiding all collisions and demonstrating the system's reliability in performing its task. For dynamic obstacles, the robot could move to the target point with an error of less than 6%, with collisions occurring in 2 out of 10 trials. The average failure to avoid collisions was primarily due to poor obstacle detection by the image processing system