{ "background": "Transport maintenance depots are critical infrastructure for road network reliability, yet systematic evaluations of their operational performance in developing contexts are scarce. Existing assessments often lack rigorous causal inference frameworks, relying on descriptive statistics that fail to isolate system-specific effects.", "purpose and objectives": "This study aimed to develop and apply a quasi-experimental methodology to quantitatively evaluate the reliability of transport maintenance depot systems, using the Ugandan context as a case study. The primary objective was to measure the causal effect of depot system interventions on equipment availability.", "methodology": "A difference-in-differences design was employed, comparing treatment and control groups of depots before and after a systemic procurement and training intervention. Reliability was operationalised as mean time between failures (MTBF) for a fleet of grader units. The core statistical model was $MTBF{it} = \\beta0 + \\beta1 (Treati \\times Postt) + \\gammai + \\deltat + \\epsilon{it}$, where $\\gammai$ and $\\deltat$ are depot and time fixed effects. Robust standard errors were clustered at the depot level.", "findings": "The intervention produced a statistically significant positive effect on system reliability. The coefficient $\\beta_1$ was estimated at 42.7 operational hours (95% CI: 18.3, 67.1), representing a 28% increase in MTBF for depots in the treatment group relative to controls.", "conclusion": "The quasi-experimental design successfully isolated the effect of targeted systemic improvements, demonstrating that structured interventions in procurement and training can substantially enhance the technical reliability of maintenance depot systems.", "recommendations": "Infrastructure agencies should adopt quasi-experimental frameworks for programme evaluation. Policy should prioritise integrated procurement-and-training packages over isolated equipment supply, and institutionalise the collection of high-frequency operational data for ongoing monitoring.", "key words": "infrastructure management, maintenance engineering, causal inference, difference-in-differences, system reliability