Condition-based maintenance for OEM’s by application of data mining and prediction techniques

Increasing the products service life and reducing the number of service visits are becoming more and more top priorities for Original Equipment Manufacturer companies (OEM’s). Condition-based maintenance is often proposed as a solution to reach this goal. However, this latter, is often hampered by the lack of the right information that gives a good indication of the health of the equipment. Furthermore, the processing power needed to compute this information is often not afforded by machine’s processor. In this paper, a remote platform which connects the OEM’s to the customer’s premises is described, allowing thus a local computation of available information. Two approaches are then combined to process the optimal maintenance time. First, data mining techniques and reliability estimation are applied to historical databases of machines running in the field in order to extract the relevant features together with their associated thresholds. Second, prediction algorithm is applied to the selected features in order to estimate the optimal time to preventively perform a maintenance action. The proposed method has been applied to a database of more than 2000 copy machines running in the field and proved to identify easily the relevant features to be forecasted and to offer an accurate prediction of the maintenance action.

[1]  David G. Stork,et al.  Pattern classification, 2nd Edition , 2000 .

[2]  Pierre Geurts,et al.  Contributions to decision tree induction: bias/variance tradeoff and time series classification , 2002 .

[3]  B. Tjahjono,et al.  A Review of Research in Manufacturing Prognostics , 2006, 2006 4th IEEE International Conference on Industrial Informatics.

[4]  Sholom M. Weiss,et al.  Predictive data mining - a practical guide , 1997 .

[5]  Everette S. Gardner,et al.  Exponential smoothing: The state of the art , 1985 .

[6]  David G. Stork,et al.  Pattern Classification , 1973 .

[7]  R. Keith Mobley,et al.  An introduction to predictive maintenance , 1989 .

[8]  Yan Zhan,et al.  An Optimization Algorithm of K-NN Classification , 2006, 2006 International Conference on Machine Learning and Cybernetics.

[9]  R. Meyer,et al.  The Fundamental Theorem of Exponential Smoothing , 1961 .

[10]  E. S. Gardner.  EXPONENTIAL SMOOTHING: THE STATE OF THE ART, PART II , 2006 .

[11]  Sigurdur Olafsson,et al.  Multi-Attribute Decision Trees and Decision Rules , 2006 .

[12]  J. Blair,et al.  Diagnosis and prognosis of bearings using data mining and numerical visualization techniques , 2001, Proceedings of the 33rd Southeastern Symposium on System Theory (Cat. No.01EX460).

[13]  Om Prakash Yadav,et al.  Complex system reliability estimation methodology in the absence of failure data , 2008, Qual. Reliab. Eng. Int..

[14]  Rolf Isermann,et al.  Fault-Diagnosis Systems , 2005 .

[15]  Chein-I Chang,et al.  Neural network technology for stock market index prediction , 1994, Proceedings of ICSIPNN '94. International Conference on Speech, Image Processing and Neural Networks.