Abstract
{ "background": "Industrial machinery fleets in developing economies are critical for productivity, yet systematic analysis of their operational yield is limited. In Uganda, a lack of robust methodological frameworks hinders the empirical measurement of performance improvements and the identification of key drivers.", "purpose and objectives": "This working paper aims to methodologically evaluate approaches for analysing fleet systems and to develop a panel-data estimation model for quantifying yield improvement. The objective is to provide a replicable analytical framework for engineering management.", "methodology": "We construct a novel unbalanced panel dataset from maintenance and operational logs of heterogeneous machinery across multiple industrial sites. The core econometric specification is a two-way fixed effects model: $Y{it} = \\alpha + \\beta X{it} + \\mui + \\lambdat + \\epsilon{it}$, where $Y{it}$ is the availability-adjusted yield. Inference is based on cluster-robust standard errors to account for within-fleet serial correlation.", "findings": "The methodological evaluation identifies maintenance scheduling consistency as a superior predictor of yield compared to fleet age alone. The panel estimation reveals a positive and statistically significant relationship, with a one-standard-deviation improvement in preventive maintenance adherence associated with an approximate 7.5% increase in mean yield (95% CI: 5.1% to 9.9%).", "conclusion": "The proposed panel-data model provides a validated methodological framework for yield analysis in industrial machinery contexts. It demonstrates that operational practices, notably systematic maintenance, are quantifiable and significant levers for performance enhancement.", "recommendations": "Fleet managers should prioritise the implementation of data-tracking systems to enable panel analysis. Policy should support the development of standardised performance metrics aligned with the model's variables to facilitate benchmarking across sectors.", "key words": "panel data, fixed effects, yield, machinery, maintenance, operational efficiency, industrial engineering", "contribution statement": "This paper provides a novel application of panel-data econometrics to the analysis of industrial machinery fle