Robotic system for high-precision inline measurements

Abstract

Inline measurement systems are crucial for achieving reliable and 100 % quality controlin high-tech manufacturing, as they enable real-time process and parameter optimization.These systems are designed to measure various aspects of the product, including dimensions,geometries, and gap sizes, ensuring consistent quality across different product features. Inthis relation, robotic measurement systems are of particular interest due to their flexibility and ability to position a high-precision measurement tool in different orientations at multiple measurement spots.However, the performance of robotic measurement systems is limited by the inherent in accuracies of industrial robots (IRs) and environmental vibrations present in an industrial manufacturing plant. Both of these factors cause disturbance-induced relative motion with amplitudes of several tens of micrometers between the measurement tool and the sample,which can corrupt the measurement results. When aiming for precise measurements on moving objects, the challenge increases, as the sample is being moved by the conveyor system.This process-induced relative motion further reduces the achievable inline measurement precision.One key advantage of inline measurements is that there is no need to halt production by bringing goods out of the production line for quality checks, streamlining the process and avoiding costly downtime. Although inline measurements eliminate the need to halt production for quality checks, streamlining the process and avoiding costly downtime, the related process-induced relative motion reduces the achievable measurement precision.While current inline measurement solutions offer moderate precision and suffce for many of today’s applications, they are not suitable for meeting the growing precision demands of future industrial manufacturing. This thesis aims to design and integrate a robotic systemcapable of performing 3D measurements directly on moving samples with sub-micrometer precision. Facing the challenges of process- and disturbance-induced relative motion, an active sample-tracking approach is pursued. Each system component is therefore diligently designed and tailored towards the targeted application for achieving maximum system performance,which implies crucial trade-offs in the holistic system concept.A high-precision scanning confocal chromatic sensor (SCCS) is proposed, which is tailored 3 for integration in an electromagnetically actuated and free-floating measurement platform(MP) mounted to an IR. The compact SCCS shows a mass below 300 g and is capable of performing 3D measurements with an axial resolution better than 100 nm, while providing frame rates up to 1 fps. Using voice coil actuators (VCAs) for quasi-zero stiffness actuation and mechanical decoupling from the IR, orientation-independent operability is enabled, such as required for the inspection of freeform surfaces. A tracking sensor system is included in the MP design to measure the position of the SCCS relative to the sample. By means of a 6-degree of freedom (DoF) feedback controller with 450 Hz positioning bandwidth, this position is maintained constant, generating a contactless stiff link between tool and sample for the duration of the measurement. In this way, relative motion is reduced to a minimum,and laboratory-like conditions are established directly in vibration-prone environments.To further enable the sample-tracking of moving samples, the long-stroke but coarse positioning capability of the IR is combined with the high-precision short-stroke positioning ofthe sample-tracking MP using dual stage control concepts. Through this method, the IR is repositioned in such a way that the MP is maintained within its limited actuation range. Several approaches for further improvement of the overall system performance are presented,including, among others, the extension of the system’s lateral measurement area and the post-correction of the residual sample-tracking error.Experiments on samples moved by a conveyor system reveal a residual sample-tracking error of less than 500 nm, paving the way for the SCCS to perform 3D measurements with an axial precision on the sub-micrometer scale. The novel robotic measurement system consequently achieves a level of precision in motion, which has been, up to this point, only attainable instatic measurement setups within controlled laboratory environments.