Self-adaptation based on error measurements and compensation
Manufacturing of small batches of mechanical components is often plagued by inefficiencies, and waste of raw materials and energy. Process parameters are largely determined on a trial-and-error basis and production time is significantly increased by incorporation of Quality Control - Production loops. Scientists initiated the EU-funded project 'Self optimising measuring machine tools' (Sommact) to develop novel production hardware and controlling software leading to improved product quality at competitive costs. Sommact exploits three innovations. First, scientists are employing a new metrological concept to measure and monitor machine tool errors and inspect machined parts. The measuring capabilities are enhanced to the point of being a coordinate measuring machine (CMM) to provide workplace traceable measurements and input valuable data to a self-learning component. Enhanced sensor systems provide detailed information about the geometric errors of each machine motion. The error information is combined with environmental and load conditions by an adaptive control system that changes the machine process accordingly. Lastly, a self-learning algorithm accumulates knowledge of machine behaviour under varying conditions and its relationship to part quality. All project objectives have been achieved in a timely manner. Experiments demonstrated that the Sommact system can reduce errors affecting machine accuracy by more than 75 %. In addition, on-machine inspection capabilities previously only available on CMMs have been included and traceability ensured by novel calibrated reference artefacts. Sommact has provided a valuable toolkit for measurement of machine error motion with six degrees-of-freedom (DoF) in stark contrast to previous state of the art consisting of a one DoF position sensor system. Almost all of the toolkit can be used with existing machine tools. Updating of volumetric compensation tables is done in quasi real-time subject to an operator's input. Standardisation of compensation data within ISO (International Organization for Standardization) is underway. Finally, decision-making support activities include applying immediate action, requesting recalibration of the machine if it is outside of previously learned scenarios, and predicting maintenance needs. The Sommact system should facilitate important improvements in machine tool control and product quality. Sommact has been recommended as a Flagship Project and a Success Story for both the machine tools industrial sector and the European Commission's Directorate-General for Research and Innovation (RTD).
