Home // SIMUL 2020, The Twelfth International Conference on Advances in System Simulation // View article

RUL Prediction for Cold Forming Production Tooling

Authors:
Wim De Mulder
Haije Zijlstra
Alessandro Di Bucchianico

Keywords: Remaining useful lifetime; Bootstrapping; Change points; Cold forming production tooling.

Abstract:
This paper presents a work-in-progress contribution that involves a collaboration with Philips, where the goal is to predict the remaining useful lifetime for cold forming production tooling. As the data set is complex, with many outliers and missing data points, we plan to integrate multiple techniques to reliably predict the time until maintenance is required, at least including machine learning methods, bootstrapping and change point detection techniques. The latter two methods are seldomly employed in the domain of remaining useful lifetime prediction, although they deliver very useful additional information compared to mainstream prediction techniques. Despite the fact that times of failure are currently lacking, we were able to perform a useful preliminary data analysis, which resulted in the extraction of several features to be used later as input variables for RUL prediction, and we obtained an interesting unsupervised clustering of a set of selected production runs.

Pages: 54 to 58

Copyright: Copyright (c) IARIA, 2020

Publication date: October 18, 2020

Published in: conference

ISSN: 2308-4537

ISBN: 978-1-61208-831-0

Location: Porto, Portugal

Dates: from October 18, 2020 to October 22, 2020