- since 1996 -
|NDT.net Issue - 2019-11 - NEWS ||NDT.net Issue: 2019-11|
Publication: e-Journal of Nondestructive Testing (NDT) ISSN 1435-4934 (NDT.net Journal)
TU Ilmenau and Fraunhofer IZFP develop intelligent sensor technology for processing Big DataFraunhofer-Institute for Non-Destructive Testing (IZFP)476, Saarbrücken, Germany
The Ilmenau University of Technology is intensifying its research efforts in order to master the rapidly growing amount of data in the course of digitization. The novel idea for ”intelligent” sensors: Record required data for the respective work process at an early stage and thus reduce the quantity to be processed. In order to achieve the ambitious research objectives, TU Ilmenau is planning a new professorship in collaboration with the Fraunhofer-Gesellschaft, which in turn will intensify the activities of its established research group SigMaSense. The Free State of Thuringia is supporting these activities this year with 150,000 euros. On 21 October 2019, Thuringia's Science Minister Wolfgang Tiefensee at TU Ilmenau also announced further support for the expansion of the research group and its establishment as a branch office of the Fraunhofer Institute for Nondestructive Testing IZFP in Saarbrücken with a total of up to 2.4 million euros from 2021.
Giovanni Del Galdo and Florian Römer from ”SigMaSense”, Randolf Hanke (Fraunhofer IZFP), Rector Peter Scharff and Minister Wolfgang Tiefensee (from left) present the new perspectives for Ilmenau. © TU Ilmenau
Conventional data processing methods capture the entire amount of data, often resulting in highly redundant measurement signals that then have to be evaluated in a further step. This not only calls for high expenditure, but in most cases only a small portion of the accumulated data material is needed for the respective work process. Since the beginning of 2018, TU Ilmenau‘s and Fraunhofer IZFP‘s research group ”SigMaSense”(Signal Processing for Material Data Acquisition using Smart Sensors) has been working on novel methods for obtaining intelligent data. The new approach: Let the sensor only record what is actually needed. Dr. Florian Römer, scientific head of SigMaSense, intends to develop ”intelligent” sensors: ”Instead of simply recording data all the time, the sensors should learn to think and independently decide which data is relevant and which is not. This way we keep only the part of the data that is actually needed.”
Instead of large piles of data, which have to be stored and scanned for certain patterns, selected, condensed data could be obtained in this way. Welcome side effect: The new process reduces the energy consumption of the sensors, the communication interfaces and the data storage devices. The result will be not only intelligent but also ”green” sensors.
Cognitive Sensor Systems – Efficient Processes
The current research focus relates to the development of sensors that are suited to capture production-related microstructural patterns and to merge them in the sense of an individual fingerprint, a so-called ”product DNA”. In the future, these ”digital product files” will open up completely new approaches for the optimization of the individual sectors of a material and product lifecycle.