Autonomous navigation of an indoor tour guide robot

Abstract

This paper develops methodologies and techniques for autonomous navigation of a tour-guide robot with a human-robot interaction system. The designed navigation system includes global localization, dynamic path planning, local goal-seeking, safe obstacle-avoidance, behavior fusion, and autonomous robot control. The RFID module for global pose initialization is presented based on the RSSI measurements, the proposed calibration method and least square method. With the gross initial robot pose, accurate robot pose tracking are achieved by fusing the RFID data and laser scanning measurements utilizing an extended Kalman filter. The global path is generated by dynamic programming method. The Petri-net model is employed to construct the event-driven logic control sequence from the present states of the robot and assigned tour missions. An operational scenario is used to show the effectiveness of the proposed autonomous navigation method.