Seeker: Autonomous maze-navigating and ball-potting robot

This paper presents the design and fabrication of autonomous maze-navigating and ball-potting robot for National Engineering Robotics Competition (NERC), 2012. The robot navigates through a given maze; determines the color of the boxes present at predefined locations throughout the maze and pot golf balls of respective colors in them. For navigating through the maze, without colliding with the walls, a positional Integral-controller has been designed by root locus method. In order to be efficiently implementable in a low-cost PIC18f4620 microcontroller, the robot's navigation algorithm primarily builds on the technique of linearization of Sharp IR sensor. The robot then uses this information in its novel granular control algorithm, effectively maintaining the robot in the center of the corridor throughout its motion. A self-made color sensor with discrete components helps in differentiating between the color of the boxes. Experimental results show that the robot executes its tasks at a speed of 70cm/s without experiencing any collision with the walls. Such robots find their application in industrial pick-and-place tasks.