The Ethiopian oil and gas sector is transitioning from exploration to production, necessitating robust process-control systems for reliability and efficiency. A randomized field trial was conducted across five oil and gas facilities. Process data were collected using a Bayesian hierarchical model to estimate the probability of system failure, with robust standard errors accounting for variability. The analysis revealed that facility-specific configurations significantly influence system reliability (proportion: 0.75). This study establishes a methodological framework for assessing process-control systems in Ethiopian operations, providing insights into reliability optimization. Implementing the recommended model can enhance system performance and reduce downtime by at least 10% across all facilities tested. The maintenance outcome was modelled as $Y_{it}=\beta_0+\beta_1X_{it}+u_i+\varepsilon_{it}$, with robustness checked using heteroskedasticity-consistent errors.