Abstract
{ "background": "The management of industrial machinery fleets represents a significant capital and operational expenditure for heavy industries. Current cost-effectiveness evaluations often rely on deterministic models that fail to adequately account for operational variability and uncertainty inherent in complex, multi-site systems.", "purpose and objectives": "This short report presents a novel Bayesian hierarchical modelling framework designed to quantify the cost-effectiveness of industrial machinery fleets. The objective is to provide a robust probabilistic method for integrating heterogeneous operational data to inform maintenance and replacement decisions.", "methodology": "A Bayesian hierarchical model was developed, specified as $y{ij} \\sim \\text{Normal}(\\mu{ij}, \\sigma^2)$, $\\mu{ij} = \\alphai + \\beta X{ij}$, $\\alphai \\sim \\text{Normal}(\\mu{\\alpha}, \\tau^2)$, where $y{ij}$ is a cost-effectiveness metric for machine $j$ in fleet $i$. The model incorporates operational hours, maintenance costs, and downtime data. Inference was performed using Markov chain Monte Carlo sampling.", "findings": "The model application to a case study fleet indicated a high posterior probability (exceeding 0.85) that a predictive maintenance strategy was more cost-effective than a reactive one. A key theme was the substantial fleet heterogeneity, with the group-level variance parameter $\\tau^2$ indicating that nearly 30% of the variation in cost-effectiveness was attributable to differences in fleet management practices.", "conclusion": "The Bayesian hierarchical framework provides a statistically rigorous tool for assessing machinery fleet performance, explicitly modelling uncertainty and variability across organisational units. It moves beyond point estimates to a full probabilistic evaluation.", "recommendations": "Industry practitioners should adopt probabilistic modelling to capture uncertainty in lifecycle cost analyses. Further research should integrate real-time sensor data into the model's observational layer to enhance predictive capability.", "key words": "Bayesian inference, lifecycle costing, maintenance optimisation, heavy equipment, probabilistic modelling", "contribution statement": "