Monitoring fractional nonconformance for short-run production

ABSTRACT

Quality characteristics observed in industrial processes are not always free from measurement errors. The term fractional nonconformance refers to the probability of an error-prone observation breaching the specification limits. Four new control statistics based on the fractional nonconformance concept are defined for process monitoring purposes. This work, motivated by milk products manufacturing, is tailored for short-run productions in which only individual measurements are accumulated over time. The performance of the newly defined control statistics is evaluated using simulation for both independent and autocorrelated processes. The results show that fractional nonconformance charts can be useful to monitor short-run production process, and the choice of the monitoring scheme does not heavily depend on the distribution of the quality characteristics.

KEYWORDS:

• autocorrelation
• control chart
• fractional nonconformance
• short-run production
• statistics

About the authors

Xin Zhou is a PhD student under Dr. Kondaswamy Govindaraju's supervision at Massey University, New Zealand. He obtained his Master degree in Statistics from National University of Singapore and Bachelor degree in Mathematics from Dalian University of Technology, China. His research interests are statistical quality control, statistical process control and acceptance sampling.

Kondaswamy Govindaraju is a Senior Lecturer in statistics in the Institute of Fundamental Sciences at Massey University, New Zealand. His research interests include acceptance sampling, SPC, DoE, and quality management.

Geoff Jones is an Associate Professor of Statistics in the Institute of Fundamental Sciences at Massey University, New Zealand. His main research interests are Biostatistics and Bayesian analysis.

Acknowledgments

Thanks to Roger Kissling of Fonterra for his guidance and suggestions. Detailed and helpful comments of the reviewers are also gratefully acknowledged.

Funding

This research is supported by the Primary Growth Partnership (PGP) scheme sponsored by Fonterra Cooperative Group and the New Zealand Government.