A novel robot-assisted calibration procedure for Optical Coordinate Measuring Systems

Metrology systems in general require calibration and verification procedures to provide their user with confidence in measured results. Large Volume Metrology (LVM) applications pose a unique challenge in this regard due to their large scale and ever-increasing accuracy requirements. Current methods for calibration in LVM generally require either a cumbersome artefact, costly instrumentation, or a time-consuming procedure which limits their applicability. Presented in this article is a calibration procedure for Optical Coordinate Measuring Systems (OCMSs) for use within LVM applications. The proposed approach enables the assessment of point-to-point accuracy of a system by triangulation of a single, low-cost ball bar artefact, removing the need for a high-accuracy reference measurement system. A robotic system is used to manipulate the artefact to positions throughout the measurement volume, and by connecting virtual points in 3D space with an artefact of known length, a mesh with fully defined edge lengths is formed. Relating the measured positions of each point in this mesh with the nominal positions, the point-to-point errors of the measurement system are calculated. The capability of this approach is demonstrated with a numerical simulation, where it is shown to be capable of calibrating an arbitrarily large volume, with residual errors converging to the random noise of the system. Finally, an implementation is shown using a real-world commercial Optical Motion Capture system, where, by applying a correction according to the identified point-to-point errors, the mean error of a 1.2x1.0x1.5 m measurement volume was reduced from 0.601 mm to 0.379 mm. A thorough search of the relevant literature indicates that this is the first attempt to use a single ball bar artefact and robotic system for the systematic point-to-point calibration of a measurement system. • A novel procedure for calibration of Optical Coordinate Measuring Systems (OCMS). • Based on the triangulation of a single ball bar artefact for point-to-point calibration. • Calibrates error up to the limit of random noise across a variety of error profiles. • Identifies the magnitude of systematic errors in a real-world system to within 5.9%.