Tool Center Point Calibration Method for an Industrial Robots based on Spheres Fitting Method

Abstract

As robots become more sophisticated, they can handle an increasing number of industrial processes. Six-degree-of-freedom (6DoF) robots are capable of complex motions that allow them to perform many complicated industrial tasks such as palletizing, handling, gluing, and welding using a wide variety of tools. However, to take advantage of the capabilities of these robots, accurate recalibration is required each time a new tool is introduced. There are various methods for calibrating the tool center point (TCP). Contact with reference parts, the use of distance sensors, and the use of laser interferometry are just a few. These methods show acceptable results for some industrial applications with maximum position error in the range of 2mm and 0.8mm. External sensors, such as camera systems, can also be attached to different locations on the robot to acquire the precise position of a reference object from which the robot can be calibrated. These non-contact calibration methods are more accurate and efficient than the traditional multi-point contact method, but the complex system components and high cost limit their application. The goal of this work is to improve accuracy of results given by ‘four-point calibration’ method provided by the industrial robot. A sphere fitting calibration approach based on the four-point calibration method findings as initial data is proposed in this study to avoid the exact point-to-point matching operation in the low-cost multi-point calibration method.