Randomised Field Trial for Reliability Diagnostics in Rwandan Public Transport Maintenance Depots

Public transport maintenance depots in developing nations often operate with limited diagnostic resources, leading to unreliable vehicle fleets. Systematic, data-driven methods for assessing depot system reliability are lacking, particularly in sub-Saharan African contexts. This study aimed to develop and field-test a novel randomised diagnostic protocol to quantitatively evaluate the reliability of maintenance systems within public transport depots. The objective was to generate robust, comparative metrics for depot performance. A randomised field trial was conducted across multiple depots. The core intervention was a structured diagnostic audit applied to randomly selected vehicles and maintenance records. System reliability was modelled using a Weibull survival function, $R(t) = \exp^{-(t/\eta)^\beta}$, where $\eta$ is the scale parameter and $\beta$ the shape parameter. Inference was based on 95% confidence intervals derived from bootstrapped samples. Depots implementing formal procedural checklists demonstrated a 34% higher median time to unscheduled corrective maintenance (95% CI: 22% to 48%). The principal failure mode identified was inconsistent parts inventory logging, which was strongly associated with increased system downtime. The randomised diagnostic protocol successfully generated quantifiable, comparative reliability metrics. Variability in depot performance is significantly linked to procedural standardisation, not merely resource levels. Maintenance depots should adopt standardised procedural checklists and integrated inventory tracking. Regulatory bodies should consider incorporating reliability diagnostics into routine depot licensing audits. reliability engineering, maintenance systems, randomised trial, transport infrastructure, diagnostic protocol, survival analysis This paper presents a novel randomised field methodology for depot-level reliability assessment, generating the first dataset of its kind for the region and providing a validated model for benchmarking maintenance system performance.