Full fart framåt! : Testar hastighetsgränsen för hinderunvikande med hjälp av vektorfältshistogram

The purpose of the study is to observe the dynamics of robot obstacle avoidance using Robot Operating System (ROS) and to compare the outcome using Vector Field Histogram (VFH) algorithm when incrementally increasing the speed of the simulated robot. TurtleBot3 and Gazebo are used to simulate the robot, in order to test the limits of the algorithm and minimize the potential time spent on data transfers between physical camera to analysis, to algorithm, to robot. The results show that a small increase in speed “smoothens” the movements of the robot by overpowering the micro-adjustments that are made at slower speeds. While VFH is effective at low to moderate speeds, its ability to maintain stability and obstacle avoidance diminishes as velocity increases beyond a doubling in speed (for the implementation used). At high speeds, the combination of perceptual lag, reactive overshoot, and physical inertia leads to failure.