Towards predictive quality management in assembly systems with low quality low quantity data – a methodological approach


Loading...

Date

2019

Publication Type

Conference Paper

ETH Bibliography

yes

Citations

Altmetric

Data

Abstract

Intervention costs in machine assembly increase rapidly with assembly progress. For early interventions, multivariate control charts or predictive quality management systems can be installed, yet they require large and high-quality datasets. In discrete manufacturing, data is limited by the quantity produced, making it cumbersome to obtain the required quantity for statistical modeling. In this study, a methodology for the setup of predictive quality management systems is presented. It demonstrates strategies for low-quality low-quantity datasets in discrete production. The boundary conditions of the assembly system, the process requirements, and the combination of physical and statistical modeling via feature engineering are highlighted.

Publication status

published

Book title

12th CIRP Conference on Intelligent Computation in Manufacturing Engineering

Journal / series

Volume

79

Pages / Article No.

125 - 130

Publisher

Elsevier

Event

12th CIRP Conference on Intelligent Computation in Manufacturing Engineering (CIRP ICME 2018)

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Predictive Quality Management; Assembly Systems; Machine tools; Advanced Quality Management; Statistical Modeling for Quality Management; Low-Volume Production; Combination of Physical and Statistical Data

Organisational unit

03641 - Wegener, Konrad (emeritus) / Wegener, Konrad (emeritus) check_circle

Notes

Funding

Related publications and datasets