Manufacturing systems in Tanzanian plants are often complex and variable, necessitating robust methods to assess their reliability. A Bayesian hierarchical model was applied to data from several Tanzanian plants. This approach accounts for variability among different plants and within the same plant over time. The model successfully captured systematic differences in reliability performance between plants, with significant variations observed across sites. The Bayesian hierarchical method proved effective in quantifying system reliability, offering insights into manufacturing process improvements. Future studies should explore broader datasets to validate the model's applicability and generalizability. The maintenance outcome was modelled as $Y_{it}=\beta_0+\beta_1X_{it}+u_i+\varepsilon_{it}$, with robustness checked using heteroskedasticity-consistent errors.