Bayesian Hierarchical Model for Yield Improvement in Process-Control Systems: A South African Perspective

Process-control systems are critical for optimising yield in manufacturing processes across various industries. In South Africa, these systems can enhance efficiency and reduce waste. A BHM was employed to analyse data from multiple sites, accounting for variability across different conditions and processes. The model incorporates uncertainty quantification through robust standard errors. The analysis revealed that incorporating site-specific parameters significantly improved the accuracy of yield predictions compared to a generic approach. The BHM demonstrated its effectiveness in enhancing yield measurement precision, offering insights for process-control system optimization in South Africa. Adoption of the BHM should be encouraged as it provides more nuanced and site-specific yield measurements, aiding in decision-making processes. process control, Bayesian hierarchical model, yield improvement, manufacturing systems, South Africa The maintenance outcome was modelled as $Y_{it}=\beta_0+\beta_1X_{it}+u_i+\varepsilon_{it}$, with robustness checked using heteroskedasticity-consistent errors.