%0 Journal Article
%T Hardware-Efficient System State Detection for Embedded Condition Monitoring
%V 10
%N 5
%P 1-4
%W http://hdl.handle.net/20.500.12708/228706
%U https://ieeexplore.ieee.org/document/11478451
%X Condition monitoring of devices and control systems is essential for predictive maintenance and can lower maintenance costs and downtimes significantly. Many approaches rely on machine learning methods, which, however, require substantial computational resources. If condition monitoring is to be embedded directly on a sensor or small edge device, then efficient hardware-based solutions are needed. This letter presents a minimized hardware implementation of a condition monitoring algorithm to allow resource-saving integration and a further reduction of computational and communication overhead. We compare the optimized hardware version with a more traditional implementation of the algorithm. Furthermore, a tradeoff analysis examines the effect of sample bit widths on gate count and accuracy. For the selected case study, the gate count has been reduced by 55% to 2120 gates and the needed memory size by roughly 86% to 3092 bits while still yielding system state detection results at the same level of accuracy.
%J IEEE Sensors Letters
%A Lehninger, Philipp
%A Elahi, Ardavan
%A Schnöll, Daniel
%A Jantsch, Axel
%A Sauter, Thilo
%D 2026-05
%K Circuit analysis
Circuits
Circuits and systems
Field programmable gate arrays
Filtering
Filters
Frequency modulation
Integrated circuits
Logic gates
Radio broadcasting
Sensor signal processing
condition monitoring
context awareness
publication
sensor data processing