•  
  •  
 

Abstract

To allow service robots to patrol indoors is a difficult challenge of the robot control technology. As long as this purpose can be achieved, it can increase the value of service robots in the commercial applications. In this research, an anchor deployment strategy combined with a stereo vision system for the robot synchronization positioning and mapping is proposed, which is based on an open source Gmapping Simultaneous Localization and Mapping (SLAM) algorithm of Robot Operating System (ROS), and enables a robot to simultaneously positioning and build environment maps autonomously. The anchors are developed based on the radio frequency identification (RFID), which can make the robot build the map with a shorter path independently. The proposed system architecture can also reduce the built map error caused by manipulating the robot. Furthermore, the shorter path of the robot moving also reduces the error accumulated by the odometer, and the better quality of the map is obtained. Finally, a built map is combined with the A star path planning algorithm to achieve the autonomous patrolling task. Several experimental results show that maps constructed by anchors have better mapping effects, the robot reaches a target point more accurately, and improve the success rate of sign-in.

COinS