Kinematic Identification of Industrial Robot Using End-Effector Mounted Monocular Camera Bypassing Measurement of 3-D Pose

Abstract

This article proposes a new method for kinematic identification of an industrial robot using a monocular camera mounted on its end-effector. Measurement of end-effector 3-D poses required as input for this process is bypassed by measurements solely in 2-D image space, thus permitting estimation of corrections in parameters in a single stage. This is in contrast with multiple stages used earlier. The use of camera enabled the combination of geometric and parametric techniques of identification. The advantages are well-defined experimental strategy and identification of externally defined coordinate frames. Kinematic identification using measurements in 2-D image space was not affected by issues associated with observability which is commonly encountered when measurements of 3-D pose are used. This is because of the fact that the algorithm used for this purpose utilized identification Jacobian which relates measurements in image space to corrections in parameters. It was computed using backward recursion. Hand-eye transformation was also obtained during the identification process, thus minimizing the number of steps or hardware that is involved. The results were verified through simulation and validated through multiple experiments including those based on ISO standards, on KUKA KR 5 arc industrial robot. Some of the results were comparable to that obtained using a laser tracker which is a popular but expensive and offline measurement system which is not commonly available. Robustness of the proposed method to measurement noise has also been demonstrated through simulation. The experiments were conducted and the estimated parameters could be used in an already existing system used for bin-picking.