Motion control of a cable-suspended robot using image recognition with coordinate transformation

This study will concern the implementation of a vision-based controller for cable-suspended robots, and especially for the pick-and-place function. This work will develop a novel algorithm that combines cable-suspended robot coordinate transformation with image recognition to manipulate the suspended gripper to the desired position for material-handling tasks. Two webcams sense the position of the end-effector; one will be used to calculate the horizontal planar coordinate of the target, and the other will be used to determine the vertical height of the target. Accordingly, the two-dimensional spatial image can be converted into three-dimensional image coordinates. Based on the sensed information, the robot can convert the position of the target to required cable lengths and then drive the servomotor to retract (or release) the cables to move the suspended gripper to the desired position. Various experiments will be conducted to verify the proposed methodology. The framework that is developed herein is easily implemented for real-time control and is suitable for inclusion in mechatronic kits in a university control course. The large variety of ideas that are addressed in this work apply to cranes, aerostats, and other large-scale manufacturing equipment that requires cable-driven robots.