A Comprehensive Review of Statistical Process Control Techniques and Their Applications

The literature reviewed in this paper shows Statistical Process Control developing not as a single fixed method but as a continuing sequence of responses to specific limitations exposed by earlier work. Shewhart's original chart established the conceptual distinction between common-cause and special-cause variation but was slow to detect small, sustained shifts; CUSUM and EWMA were developed specifically to address that limitation, and a subsequent wave of research refined their initial response, run-length behaviour, and parameter selection. As monitoring problems grew to involve multiple correlated characteristics, the same memory-based logic was extended into the multivariate domain, generating its own diagnostic challenge identifying which variable was responsible for a signal which later decomposition and regression-adjusted methods addressed. In parallel, the normality assumption underlying nearly all of these methods motivated a distinct nonparametric line of development, extended in turn to the multivariate case, while economic-design and adaptive-sampling research addressed the practical question of how and when to sample rather than what statistic to plot. Across this trajectory, traditional univariate charts remain the appropriate choice where interpretability and simplicity are paramount and shifts are expected to be moderate or large; memory-based univariate charts are preferable when small, sustained shifts matter; multivariate methods become necessary once several correlated characteristics must be jointly monitored; and nonparametric methods are warranted when normality cannot reasonably be assumed. No technique in the reviewed literature dominates across all of these considerations simultaneously, which is itself the review's central finding: the appropriate SPC method is a function of the specific monitoring problem, not a single universally superior chart. The gaps that remain high-dimensional and streaming data, real-time and Industry 4.0-enabled monitoring, and the integration of machine learning with classical SPC theory are not a disconnected list of fashionable topics but a direct continuation of the same developmental logic traced throughout this review: each responds to a limitation of the methods that currently exist, in much the same way that CUSUM responded to the limitations of the Shewhart chart nearly a century ago. Future research that extends the reviewed literature along these lines, while preserving the statistical rigour that has characterised the field since Shewhart's original work, is likely to determine how effectively Statistical Process Control adapts to the increasingly complex and automated production environments it is now being asked to monitor.

1.      Bersimis, S., Psarakis, S., & Panaretos, J. (2007). Multivariate statistical process control charts: An overview. Quality and Reliability Engineering International, 23(5), 517–543. 2.      Capizzi, G., & Masarotto, G. (2003). Adaptive control charts for monitoring processes. Journal of Quality Technology, 35(3), 285–298. 3.      Chakraborti, S., van der Laan, P., & Van de Wiel, M. A. (2004). A class of distribution-free control charts. Journal of the Royal Statistical Society: Series C (Applied Statistics), 53(3), 443–462. 4.      Champ, C. W., & Woodall, W. H. (1987). Exact results for Shewhart control charts with supplementary runs rules. Technometrics, 29(4), 393–399. 5.      Crosier, R. B. (1988). Multivariate generalizations of cumulative sum quality-control schemes. Technometrics, 30(3), 291–303. 6.      Crowder, S. V. (1987). A simple method for studying run-length distributions of exponentially weighted moving average charts. Technometrics, 29(4), 401–407. 7.      Duncan, A. J. (1956). The economic design of X̄ control charts used to maintain current control of a process. Journal of the American Statistical Association, 51(274), 228–242. 8.      Hawkins, D. M. (1991). Multivariate quality control based on regression-adjusted variables. Technometrics, 33(1), 61–75. 9.      Khoo, M. B. C. (2004). Moving average control charts for monitoring the fraction nonconforming. Quality Engineering, 16(3), 399–408. 10.   Knoth, S. (2005). Accurate ARL computation for EWMA control charts. Journal of Quality Technology, 37(1), 78–89.