D2.4 Lean data with edge analytics: Decentralized current profile analysis on embedded systems using neural networks
- Event
- SMSI 2020
-
(did not take place because of Covid-19 virus pandemic) - Band
- SMSI 2020 - Measurement Science
- Chapter
- D2 AI-Approaches in Measurement
- Author(s)
- T. Küfner - University Bayreuth, Bayreuth (Germany), A. Trenz - Fraunhofer IPA, Bayreuth (Germany), S. Schönig - Maxsyma GmbH & Co. KG, Floss (Germany)
- Pages
- 271 - 272
- DOI
- 10.5162/SMSI2020/D2.4
- ISBN
- 978-3-9819376-2-6
- Price
- free
Abstract
This short paper introduces a system for the detection of operating states based on current profiles of a production plant with an artificial neural network at the machine’s edge in almost real-time. The system called “CogniSense” consists of a sensor for signal acquisition, a microcontroller for data preprocessing and a single-board computer for data main processing. With the system, current profiles of a test engine are acquired and analyzed, so that 26 defined operating states can be reliably detected with a classification accuracy of over 95%.