Fuzzy Inference-Based Person-Following Robot

Abstract

Abstract—Person-following is an important ability that needs to be possessed by a service robot when required to accomplish some human-related tasks. Such ability has requirements, which cannot be met satisfactorily using classical mathematical methods. Most notably, the robot has to remain at a certain safe distance from the person that is being followed and at the same time move in a smooth manner that does not appear threatening to the person. In this paper, therefore, a fuzzy inference system is developed and used as a controller to provide decisions achieving smooth and safe person-following behavior. The Fuzzy system is made to work in conjunction with a leg detection algorithm and a laser range finder to detect a person's legs giving the inference system distance and velocity information necessary for the control process. The experimental results showed that even though the detection of legs was subject to noise and false negatives, the robot achieved the smoothness and safety objectives while following its target.