Episode rule-based prognosis applied to complex vacuum pumping systems using vibratory data

Authors:
Florent Martin;Nicolas Méger;Sylvie Galichet;Nicolas Becourt
Affiliations:
University of Savoie, Polytech'Savoie, LISTIC laboratory, Annecy-le-Vieux, France and Alcatel Vacuum technology, Annecy, France;University of Savoie, Polytech'Savoie, LISTIC laboratory, Annecy-le-Vieux, France;University of Savoie, Polytech'Savoie, LISTIC laboratory, Annecy-le-Vieux, France;Alcatel Vacuum technology, Annecy, France
Venue:
ICDM'10 Proceedings of the 10th industrial conference on Advances in data mining: applications and theoretical aspects
Year:
2010

Citing 6
Cited 0

Discovery of Frequent Episodes in Event Sequences

Data Mining and Knowledge Discovery
Data Mining to Predict Aircraft Component Replacement

IEEE Intelligent Systems
Mining Sequential Patterns

ICDE '95 Proceedings of the Eleventh International Conference on Data Engineering
Machine Learning in Medical Applications

Machine Learning and Its Applications, Advanced Lectures
Constraint-based mining of episode rules and optimal window sizes

PKDD '04 Proceedings of the 8th European Conference on Principles and Practice of Knowledge Discovery in Databases
Continuously matching episode rules for predicting future events over event streams

APWeb/WAIM'07 Proceedings of the joint 9th Asia-Pacific web and 8th international conference on web-age information management conference on Advances in data and web management

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a local pattern-based method that addresses system prognosis. It also details a successful application to complex vacuum pumping systems. More precisely, using historical vibratory data, we first model the behavior of systems by extracting a given type of episode rules, namely First Local Maximum episode rules (FLM-rules). A subset of the extracted FLM-rules is then selected in order to further predict pumping system failures in a vibratory datastream context. The results that we got for production data are very encouraging as we predict failures with a good time scale precision. We are now deploying our solution for a customer of the semi-conductor market.