Met4FoF – Metrology for the Factory of the Future

The Factory of the Future (FoF) as a networked production environment with an autonomous flow of information and decision-making represents the digital transformation of manufacturing to increase efficiency and competitiveness. Transparency, comparability and sustainable quality assurance require reliable measurement data, processing methods and results.

problem

A comprehensive metrological infrastructure includes traceable calibrations, the treatment of measurement uncertainties as well as industry standards and guidelines. Digitalization is changing almost all aspects of this infrastructure: For example, sensors become intelligent and large sensor networks are used together with ML algorithms to make automated decisions and control production processes. Data quality is one of the most important industrial needs in the “factory of the future” addressed by the Met4FoF project. Methods for measurement technology in sensor networks need to be extended and real-time ML methods developed to enable measurement uncertainty assessment in industrial sensor networks.

objective

The Metrology for the Factory of the Future (Met4FoF) project creates a metrological framework for the entire life cycle of measurement data in industrial applications: from the calibration of individual digital sensors to the determination of measurement uncertainty related to machine learning (ML) in industrial sensor networks. The implementation in realistic test benches demonstrates the practical applicability of the procedures and provides templates for further application areas and future use by the industry.

Recovery concept:

The technical work focuses on two common manufacturing challenges, process optimization and predictive maintenance, represented by three specific test benches with different types of sensor networks.
The ZeMA test bench uses a range of sensors that measure different sizes for end-of-line testing and condition monitoring of electromechanical cylinders. The measurement uncertainty is to be taken into account throughout the entire data flow, i.e. from the individual sensors to the output of data analysis using machine learning. The toolkit developed at ZeMA for feature extraction and selection as well as for the classification and validation of cyclic sensor data will be extended within the scope of the project in particular by the measurement uncertainty analysis for all methods used.

Project management: Tanja Dorst
Project management: Prof. Dr. Andreas Schütze
Duration: 01.05.2018 – 31.05.2021

This project 17IND12 Met4FoF has received funding from the EMPIR programme co-financed by the Participating States and from the European Union’s Horizon 2020 research and innovation programme.

Category: Multimodal smart sensing

Other projects