Automated increase in productivity through feedback of data from quality assurance in machining production

Key Info

Basic Information

01.02.2019 to 31.01.2021
Organizational Unit:
Chair of Machine Tools, Machine Data Analytics and NC-Technology
German Federation of Industrial Research Associations AiF

Research partner

  • Forschungsvereinigung Programmiersprachen für Fertigungseinrichtungen e. V. (FVP)


Today's market is characterized by smaller batch sizes and simultaneously increasing demands on component quality. As a result, resource-intensive work steps, such as the design and running-in of new manufacturing processes, are gaining in importance and are increasingly becoming competitive bottlenecks. Although simulation-based methods already exist for process-specific optimization, their use is often too costly and therefore uneconomical, especially for small and medium-sized enterprises - SMEs. In order to ensure compliance with quality requirements, processes are therefore usually designed conservatively with reduced cutting parameters, with the disadvantage that residual potentials regarding machine performance and component tolerances remain unused. Information about the process is essential in order to uncover optimization potentials. In addition to simulative generation, this information can also be obtained from quality measurement data. Although the latter are anyway part of the quality assurance process, the measurements are nowadays mostly only used to monitor the process, but not to optimize it. In other words, quality assurance feedback is usually only given if the required quality is not maintained, but not if tolerances and thus productivity remain untapped.

By using quality measurement data, the process information required for productivity optimization can be generated without additional measurement effort. If the evaluation of the quality data and the optimization of the process were to be automated as far as possible, this would create an attractive tool for process optimization in small batches, especially for SMEs. The AutoPRO research project presented here addresses precisely this vision. The AutoPRO approach provides for an automated iterative optimization of production parameters and/or path paths by actively returning the results from quality assurance in order to increase the productivity of production processes. The path information of the machine tools is to be linked in a software with the quality requirements from the construction and the quality measurement data. Subsequently, this software automatically evaluates the individual data sources and creates a concrete recommendation for the user regarding the production parameters and/or corresponding path adjustments.

The optimization is iterative. The first step is conservative planning, which is then optimized step-by-step over several components. By setting up an intelligent optimization algorithm, an explicit understanding of cause-effect relationships is to be achieved that will enable continuous improvement of the manufacturing processes in the future.