Abstract
{ "background": "Maintenance regimes for industrial machinery fleets in sub-Saharan Africa are often based on fixed schedules or reactive practices, with limited empirical evaluation of their effectiveness in mitigating operational risks within specific regional contexts.", "purpose and objectives": "This study aimed to methodologically evaluate the comparative efficacy of a novel risk-based maintenance (RBM) regime against conventional time-based maintenance (TBM) for industrial machinery fleets, with the primary objective of quantifying associated reductions in critical failure risk.", "methodology": "A randomised field trial was conducted using a fleet of 42 heavy-duty vehicles from three industrial sites. Units were randomly assigned to either the RBM or TBM group. The RBM intervention utilised a proportional hazards model, $\\lambda(t|X) = \\lambda0(t) \\exp(\\beta1 X1 + \\beta2 X_2)$, to prioritise tasks based on real-time risk indicators. Key performance metrics, including mean time between critical failures (MTBCF) and downtime cost, were monitored over an operational period.", "findings": "The risk-based regime reduced the incidence of critical failures by 32% (95% CI: 24% to 40%) compared to the time-based approach. Furthermore, analysis using robust standard errors indicated a statistically significant reduction in total unplanned downtime costs per vehicle for the RBM cohort.", "conclusion": "The risk-based methodology demonstrated superior performance in mitigating high-consequence failure events within the studied operational environment, proving its practical viability for asset-intensive industries in the region.", "recommendations": "Industrial operators should adopt risk-based maintenance frameworks, supported by continuous data collection on machine condition and failure history, to optimise resource allocation and improve fleet reliability.", "key words": "risk-based maintenance, field trial, machinery fleet, proportional hazards model, comparative study, asset management", "contribution statement": "This study provides the first randomised field-trial evidence for the efficacy of a formal risk-based maintenance methodology applied to