"background": "Industrial machinery fleets in developing economies are critical for infrastructure development, yet systematic methodologies for quantifying and improving their operational yield are lacking. Existing approaches often fail to account for heterogeneous machine types, site-specific conditions, and the inherent uncertainty in performance data. ", "purpose and objectives": "This study presents a methodological evaluation of a Bayesian hierarchical model designed to measure and diagnose yield improvement within industrial machinery fleets. The objective is to provide a robust statistical framework that integrates multi-level operational data to inform maintenance and deployment strategies. ", "methodology": "A Bayesian hierarchical model was developed and applied to operational data from a fleet of excavators, bulldozers, and haul trucks. The core model structure is y{ij \ (\ + \ Xij, \²), with \ \ (\\, \²), where yij is the yield metric for machine i on project j, and Xij represents covariates. Posterior distributions were estimated using Markov chain Monte Carlo sampling. ", "findings": "The model successfully identified significant heterogeneity in baseline performance (\) across machine units. A key finding was that for haul trucks, preventive maintenance interventions informed by the model's posterior predictions were associated with a yield increase of approximately 15% (95% credible interval: 11% to 19%) compared to a schedule-based approach. ", "conclusion": "The Bayesian hierarchical model provides a statistically rigorous and operationally actionable framework for yield analysis. It effectively quantifies improvement while characterising uncertainty, offering a superior alternative to aggregate fleet-level metrics. ", "recommendations": "Fleet managers should adopt hierarchical modelling techniques to disaggregate performance drivers. Further research should integrate real-time sensor data into the model's structure to enable dynamic forecasting. ", "key words": "Bayesian statistics, fleet management, operational yield, heavy equipment, probabilistic modelling, maintenance optimisation
Kwame Asante (Sun,) studied this question.