Using tablets in the vision-based control of a ball and beam test-bed

Although the onboard cameras of smart devices have been used in the monitoring and teleoperation of physical systems such as robots, their use in the vision-based feedback control of such systems remains to be fully explored. In this paper, we discuss an approach to control a ball and beam test-bed using visual feedback from a smart device with its camera pointed at the test-bed. The computation of a homography between the frames of a live video and a reference image allows the smart device to accurately estimate the state of the test-bed while facing the test-bed from any perspective. Augmented reality is incorporated in the development of an interactive user interface on the smart device that allows users to command the position of the ball on the beam by tapping their fingers at the desired location on the touchscreen. Experiments using a tablet are performed to characterize the noise of vision-based measurements and to illustrate the performance of the closed-loop control system.