A Bayesian model for approaching a human

Abstract

With the growing need for elder care, research is focusing on robotic assistance at home. Thus, robots must navigate in cluttered, domestic, indoor environments with the purpose of interacting with a person. Here we present a behaviour based navigation model enhanced with a low level decision making process that allows the robot to approach a human in such an environment. The model has been tested on simulation and the first results show the effectiveness of the Bayesian decision making process.