Figure 6 - Experimental setup The conceptual idea of the experiment is to utilise a camera as a measuring system, while moving the robot together with two additional external measuring systems attached to it: the 6DOF measuring arm and the iGPS. Figure 5 - Photograph of the (DOF measuring arm coupled to the KUKA robot The robot was moved systematically between each pair of points. The change in camera pixels recorded for a robot movement of 4mm was noted at different distances away from the camera. These values were then used to calculate the ratio of mm/pixel for the camera at different distances. The assumption is that the distance from the camera would have little effect on the ratio of mm/pixel observed due to the telecentric lens. The camera calibration curve is shown in Fig. 7. The obtained results of the X and Z axes error values recorded by all four measurement systems at each data point are shown in Figs. 8 and 9. The robot was calibrated by the robot manufacturer prior to the experiments. Therefore, the robot data was used as a reference. The results indicate that the error values recorded by the camera are the closest to the reference values. Figure 8 - Measured X-axis errors accuracy, followed by the 6DOF measuring arm and the iGPS. Figure 10 - Errors with respect to the robot data
