Robotic navigation of smooth contours

Abstract

The goal of this work is to develop a method for robotic navigation of smooth contours depending on the current and desired locations and orientations. Efficient trajectory generation is an essential capability for many autonomous mobile robots, operating in a variety of situations such as military, medical, and home environments. In this thesis, we propose a method that is based on fitting a spline curve that passes from the initial position and orientation of the robot to a goal position and orientation. The spline is continually recomputed as the robot moves through space. This yields a simple and inefficient method for robot navigation. The method has been implemented and tested in simulation using Matlab and good performance has been demonstrated. Future work should perform experiments with this method on a real robot and should introduce obstacle detection and avoidance.