DMAIC Six Sigma Methodology in Petroleum Hydrocarbon Oil Classification

Abstract

This research focuses on the use of the DMAIC method (Define, Measure, Analyze, Improve and Control) as a Six Sigma approach in studying oil spill fingerprint of samples recovered from Peninsular Malaysia and Sabah (East Malaysia). The DMAIC approach in this study was used as a way to classify oil types based on data obtained from GC-FID and GC-MS measurements. The cause-effect diagram was used to define the factors leading to the failure of the oil spill fingerprinting based on inaccurate oil type clustering. Discriminant Analysis (DA) was also applied to quantify the root-cause of the failure. An Ishikawa diagram obtained in the analysis phase identifies the potential failure causal. Principal component analysis (PCA) was applied and was successful in discriminating four clusters of oil types, namely diesel, heavy fuel oil (HFO), mixture oil lube and fuel oil (MOLFO) and waste oil (WO) with a total variance of 85.3%. In the control phase, the use of a Pareto chart indicated 100% cumulative percentage of oil type clustering with a 95% confidence level. The DMAIC approach to be effective in solving oil spill fingerprinting problems and results in quality improvement in the clustering of oil spills into the different hydrocarbon types.

Multivariate Six Sigma: A case study in an outpatient pharmaceutical care unit

“…A data-driven and efficient solution for these purposes is provided by Partial Least Squares (PLS) [12], [13] models. PLS is a latent variable-based multivariate statistical tool that has already been integrated into the Six Sigma toolkit in the industrial context [15]- [17], but not in the healthcare environment, as far as we are concerned.…”

Multivariate Six Sigma: A Case Study in Industry 4.0

“…For example, Peruchi et al [5] integrated principal component analysis (PCA) into a Six Sigma DMAIC project for assessing the measurement system, analyzing process stability and capability, as well as modeling and optimizing multivariate manufacturing processes in a hardened steel turning case involving two critical-to-quality (CTQ) characteristics. In [6], discriminant analysis and PCA were integrated into the DMAIC Six Sigma framework in order to improve the quality of oil type classification from oil spills chromatographic data.…”

Causal latent space-based models for scientific learning in Industry 4.0

“…In such context, latent variable-based multivariate statistical techniques are widely recommended as commented in Section 1.1. In the literature, there are some examples of this integration of multivariate statistical tools into the Six Sigma toolkit [13,114,115]. This chapter reinforces conclusions from previous works in the literature on how Six Sigma's DMAIC methodology can be used to achieve competitive advantages, efficient decision-making, and problem-solving capabilities within the Industry 4.0 context, by incorporating latent variable-based techniques such as Partial Least Squares (PLS) into the statistical toolkit leading to the so-called Multivariate Six Sigma [2].…”