Feature detection in multidimensional datasets of lubricated contacts

This cooperation is embedded in the Strategic Project “Data-driven discovery in tribology” and therein in the Work Package “Event detection in time-series data”. This Work Package deals with the development of algorithms, e.g., by use of Machine Learning (ML), that allow failure detection in data obtained from tribological experiments or plant operation. The Work Package makes use of existing experimental and analytical data that serve as training data for the recognition of characteristic features, e.g., for the early detection of impending failures during tribological experiments.

This cooperation is laid down in the Sub-Work Package “Feature detection in multidimensional datasets of lubricated contacts” focused on the feature detection in friction and wear properties caused by various states of lubrication in time-series and multi-dimensional data sets.

 

Coordinator: AC2T research GmbH

TU Wien team: Pia Pfeiffer, Peter Filzmoser

 

Program / Call: Comet K2

Proposal: InTribology: Tribology Intelligence – Customized Tribology for Industrial Innovations

Funding: The Austrian Research Promotion Agency (FFG)

Start: 01 June 2021, duration: 36 months

Project web page:

www.ac2t.at, opens an external URL in a new window