Artificial Intelligence-Driven Predictive Analytics for Academic Infrastructure Utilisation in Public Polytechnics of Southwest Nigeria

Abstract

Rapid growth in student enrolment has placed unprecedented pressure on the physical
infrastructure of higher education institutions, particularly lecture theatres, laboratories, offices,
and other teaching and administrative facilities. Despite increasing infrastructure demand,
planning and resource allocation in many Nigerian public polytechnics remain largely reactive,
relying on conventional decision-making approaches that often fail to optimise facility utilisation.
This study investigated the influence of data availability, artificial intelligence (AI) analytics
capability, and staff digital competence on infrastructure utilisation efficiency and evaluated the
predictive performance of selected machine learning algorithms for forecasting facility demand in
public polytechnics in Southwest Nigeria. A cross-sectional survey involving 318 academic and
administrative staff was combined with predictive modelling using institutional facility utilisation
records. The measurement model demonstrated satisfactory reliability and construct validity,
confirming the robustness of the research instrument. Structural model analysis revealed that AI
analytics capability exerted the strongest positive effect on infrastructure utilisation efficiency,
followed by data availability and staff digital competence, with the model explaining a substantial
proportion of the variance in institutional efficiency. Comparative evaluation of machine learning
algorithms further showed that Gradient Boosting and Random Forest consistently achieved the
highest prediction accuracy and the lowest forecasting errors, outperforming the alternative models
considered. These findings demonstrate that integrating artificial intelligence-driven predictive
analytics with high-quality institutional data and digitally competent personnel can substantially
enhance infrastructure planning, resource allocation, and operational efficiency. The study
provides empirical evidence supporting the adoption of data-driven decision-making frameworks
and intelligent forecasting systems as practical pathways toward smart campus development and
sustainable infrastructure management in resource-constrained higher education institutions.