Performance Evaluation of Human Comfortable Following Model for Service Robots

Human following for service robots is not only a key skill in mobile robotic applications, but also an important basis in human-robot friendly interaction. Since human behavior is relatively complex and influenced by many environmental factors, it is necessary to establish a motion prediction model considering behavior analysis. This paper proposes a following method considering human comfort for service robots to achieve a more accurate and effective following behavior with a high degree of acceptance. An anticipative human following behavior is proposed when the humans must be taken explicitly into account. The human model is constructed using relevant scientific studies about human walk and social interaction, allowing the robot to predict the human trajectory and to take preemptive action. On the basis of discussing the factors that affect the following comfort, we establish the comfortable following model and behavior utility function based on human comfort.