This study examines municipal infrastructure asset management in Ethiopia, focusing on identifying yield improvement opportunities. A Bayesian hierarchical model was applied to analyse data from municipal infrastructure assets, considering spatial and temporal variations. Robust standard errors were calculated to account for uncertainty. The analysis revealed a significant improvement (p < 0.01) in yield by implementing the proposed model compared to traditional methods. The Bayesian hierarchical model demonstrated effectiveness in improving yield measurement, providing a robust framework for future applications in Ethiopian municipal infrastructure management. Further research should explore scalability and potential integration with existing systems to maximise benefits. The maintenance outcome was modelled as $Y_{it}=\beta_0+\beta_1X_{it}+u_i+\varepsilon_{it}$, with robustness checked using heteroskedasticity-consistent errors.