No Cover Image

Journal article 137 views 17 downloads

Risk based uncertainty quantification to improve robustness of manufacturing operations / Cinzia Giannetti; Rajesh S. Ransing

Computers & Industrial Engineering, Volume: 101, Pages: 70 - 80

Swansea University Author: Ransing, Rajesh

  • giannetti2016(2).pdf

    PDF | Version of Record

    © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

    Download (1.11MB)

Abstract

The cyber-physical systems of Industry 4.0 are expected to generate vast amount of in-process data and revolutionise the way data, knowledge and wisdom is captured and reused in manufacturing industries. The goal is to increase profits by dramatically reducing the occurrence of unexpected process re...

Full description

Published in: Computers & Industrial Engineering
ISSN: 0360-8352
Published: 2016
Online Access: Check full text

URI: https://cronfa.swan.ac.uk/Record/cronfa29528
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract: The cyber-physical systems of Industry 4.0 are expected to generate vast amount of in-process data and revolutionise the way data, knowledge and wisdom is captured and reused in manufacturing industries. The goal is to increase profits by dramatically reducing the occurrence of unexpected process results and waste. ISO9001:2015 defines risk as effect of uncertainty. In the 7Epsilon context, the risk is defined as effect of uncertainty on expected results. The paper proposes a novel algorithm to embed risk based thinking in quantifying uncertainty in manufacturing operations during the tolerance synthesis process. This method uses penalty functions to mathematically represent deviation from expected results and solves the tolerance synthesis problem by proposing a quantile regression tree approach. The latter involves non parametric estimation of conditional quantiles of a response variable from in-process data and allows process engineers to discover and visualise optimal ranges that are associated with quality improvements. In order to quantify uncertainty and predict process robustness, a probabilistic approach, based on the likelihood ratio test with bootstrapping, is proposed which uses smoothed probability estimation of conditional probabilities. The mathematical formulation presented in this paper will allow organisations to extend Six Sigma process improvement principles in the Industry 4.0 context and implement the 7 steps of 7Epsilon in order to satisfy the requirements of clauses 6.1 and 7.1.6 of the ISO9001:2015 and the aerospace AS9100:2016 quality standard.
College: College of Engineering
Start Page: 70
End Page: 80