Comparative Evaluation of Process-Control System Methodologies for Efficiency Gains in South Africa: A Quasi-Experimental Design

{ "background": "Process-control systems are critical for operational efficiency in capital-intensive industries, yet there is a paucity of rigorous, field-based evaluations comparing the efficacy of different methodological approaches within the local industrial context.", "purpose and objectives": "This study aims to empirically compare the efficiency gains delivered by three distinct process-control system methodologies—model predictive control, statistical process control, and a rule-based heuristic system—within operational industrial plants.", "methodology": "A quasi-experimental design was implemented across three matched manufacturing sites. Operational data were collected pre- and post-intervention for each site, which implemented one of the three methodologies. Efficiency was measured via a normalised output-per-unit-energy metric. The treatment effect was estimated using a difference-in-differences model: $Y{it} = \\beta0 + \\beta1 \\text{Treat}i + \\beta2 \\text{Post}t + \\beta3 (\\text{Treat}i \\times \\text{Post}t) + \\epsilon{it}$, with inference based on cluster-robust standard errors.", "findings": "The model predictive control system yielded a statistically significant mean efficiency gain of 17.3% (95% CI: 14.1, 20.5), substantially outperforming the other two methodologies, which showed non-significant improvements. The primary mechanism for this gain was a reduction in energy consumption during transient operational states.", "conclusion": "The choice of process-control methodology has a material impact on operational efficiency. Model predictive control, while more complex to implement, demonstrated superior performance in this field evaluation.", "recommendations": "Industrial practitioners should prioritise advanced model-based control strategies for new system implementations. Further research should investigate the scalability and long-term maintenance costs associated with these methodologies.", "key words": "process control, quasi-experiment, efficiency, model predictive control, industrial engineering", "contribution statement": "This study provides the first field-based, comparative evidence from a