A Novel Static and Moving Obstacles Avoidance Method Using Bayesian Filtering for Service Robot

Abstract

This paper presents a novel static and moving obstacles avoidance method in indoor environment for service robot. This method uses combination of information about static obstacles on the landmark, speed and direction of people that walks as moving obstacle obtained by single camera, then calculated the probability of collision using Bayesian Filtering. In this research, 2 stopping point in landmark used for efficiency in detecting and processing obstacles. By increasing/lowering the speed of robot based on the state estimation calculated before, robot able to avoid the moving obstacles. Algorithms for static and moving obstacles avoidance proposed and simulation result implemented to service robot presented. This method very robust and successfully implemented to service robot called Srikandi II that equipped with 4 DOF arm robot in our laboratory.