"background": "Process-control systems are critical for infrastructure and industrial operations, yet rigorous methodological frameworks for evaluating their reliability in developing contexts are lacking. Existing assessments often rely on cross-sectional data, failing to account for temporal changes and confounding factors. ", "purpose and objectives": "This study aims to develop and apply a robust quasi-experimental methodology to evaluate the causal impact of modernising process-control systems on their operational reliability within a West African context. ", "methodology": "A difference-in-differences (DiD) model was employed, analysing panel data from treatment and control groups of systems before and after a major technological intervention. The primary model is specified as Y{it = \0 + \1 + \2 + \ (\) +, where Yit is the reliability index. Inference is based on cluster-robust standard errors. ", "findings": "The modernisation programme significantly improved system reliability. The DiD estimator \ was 0. 15 (95% CI: 0. 11 to 0. 19), indicating a 15-percentage-point increase in the reliability index for treated systems relative to controls. The parallel trends assumption was validated using pre-intervention data. ", "conclusion": "The application of a DiD framework provides a rigorous, causal methodology for evaluating engineering system performance, demonstrating a substantial positive effect of the technological upgrade on reliability. ", "recommendations": "Engineering assessments of system interventions should adopt quasi-experimental designs to isolate causal effects. Policymakers should prioritise investments in modern control architectures, supported by longitudinal monitoring data. ", "key words": "process control, reliability engineering, difference-in-differences, causal inference, infrastructure management", "contribution statement": "This paper provides a novel application of econometric causal inference methods to engineering system evaluation, establishing a
Sarr et al. (2007) studied this question.