Skip to Main content Skip to Navigation
Preprints, Working Papers, ...

Unsupervised manufacturing process identification using non-intrusive sensors

Abstract : Energy sustainability in the manufacturing industry faces a scalability issue. Monitoring appropriate performance indicators is essential, yet as few sensors as possible should be used, and with limited intrusiveness (software- or hardware-wise). Non-intrusive sensors are well suited to such applications, as multiple sources can be sensed at once. Recovering the desired indicators requires additional signal processing though. This paper focuses on recovering a machine’s process from sensor data in an unsupervised fashion, and unveiling which actuators are active within each operation. The proposed method is particularly well suited to mixed signals which appear as stationary in the time-frequency domain within each operation.
Document type :
Preprints, Working Papers, ...
Complete list of metadata

https://hal.archives-ouvertes.fr/hal-03768559
Contributor : Romain DELABEYE Connect in order to contact the contributor
Submitted on : Sunday, September 4, 2022 - 11:02:28 AM
Last modification on : Saturday, September 17, 2022 - 3:46:14 AM

Files

v2_postreview_CFM2022.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-03768559, version 1

Citation

Romain Delabeye, Martin Ghienne, Arkadiusz Kosecki, Jean-Luc Dion. Unsupervised manufacturing process identification using non-intrusive sensors. 2022. ⟨hal-03768559⟩

Share

Metrics

Record views

5

Files downloads

1