Journal Design Engineering Masthead
African Structural Engineering | 22 August 2018

A Quasi-Experimental Evaluation of Machinery Fleet Diagnostics and Yield Optimisation in Rwandan Industry

J, e, a, n, d, e, D, i, e, u, N, i, y, o, n, z, i, m, a, ,, M, a, r, i, e, C, l, a, i, r, e, U, w, i, m, a, n, a, ,, C, h, a, n, t, a, l, U, w, a, s, e, ,, S, a, m, u, e, l, H, a, b, i, m, a, n, a
DOI: https://doi.org/10.5281/zenodo.18970778
Predictive MaintenanceYield OptimisationQuasi-Experimental DesignSub-Saharan Africa
A difference-in-differences design isolates the causal effect of a diagnostic intervention on industrial yield.
Implementation of real-time machinery health monitoring resulted in a statistically significant mean yield increase of 7.3%.
Greatest efficiency gains were identified at the drying and sorting stages through pre-emptive process adjustments.
The study provides evidence for scalable data-driven maintenance strategies in developing industrial contexts.

Abstract

{ "background": "The operational efficiency of industrial machinery fleets in developing economies is often constrained by reactive maintenance and suboptimal process control, leading to significant yield losses. There is a paucity of structured, evidence-based evaluations of diagnostic interventions within these contexts, particularly in Sub-Saharan Africa.", "purpose and objectives": "This case study aims to methodologically evaluate the impact of a systematic diagnostic and optimisation programme on industrial yield. The primary objective is to quantify yield improvement using a quasi-experimental design, isolating the effect of the intervention from other operational variables.", "methodology": "A quasi-experimental, difference-in-differences design was implemented across two comparable tea processing factories. The treatment facility received an integrated diagnostic system for real-time machinery health monitoring and process parameter optimisation, while the control facility continued with its existing practices. The core impact was estimated using a fixed-effects panel model: $Y{it} = \\beta0 + \\beta1 (\\text{Treat}i \\times \\text{Post}t) + \\alphai + \\gammat + \\epsilon{it}$, where robust standard errors were clustered at the factory level.", "findings": "The intervention yielded a statistically significant positive effect. Analysis indicates a mean yield increase of 7.3 percentage points in the treatment factory relative to the control (95% CI: 5.1 to 9.5). The most substantial gains were observed in the drying and sorting stages, where diagnostic alerts enabled pre-emptive adjustments to machine settings.", "conclusion": "The structured application of fleet diagnostics and yield optimisation systems can deliver substantial operational improvements in Rwandan industrial settings. The quasi-experimental design provides credible evidence of a causal relationship between the implemented system and enhanced production efficiency.", "recommendations": "Industrial operators should consider adopting integrated diagnostic systems, prioritising stages with high yield variability. Policymakers and industry associations are encouraged to support the development of local technical capacity for sustaining such data-driven maintenance and optimisation programmes.", "key words":