ORCID
Dear Editor-in-Chief,
Thank you for the guidance in preparing the final manuscript for publication in the Tanzania Journal of Engineering and Technology (TJET). The manuscript titled “Artificial Neural Network-Based Modelling and Prediction of Key Performance Indicators in Road Construction Projects” has been carefully reviewed to address minor corrections suggested by Reviewer No. 3. Should there be any aspects requiring further revision, it would greatly be appreciated for your guidance.
Thank you very much for your guidance.
Abstract
Road construction projects in developing countries including Tanzania experience several challenges including cost overruns, schedule delays and poor quality. The causes of these challenges include complex interdependencies project’s uncertainty factors. This makes traditional methods for predicting Key Performance Indicators (KPIs) less effective. This study has developed a robust Artificial Neural Network (ANN) model for accurate modelling and predicting KPIs for road construction projects in Tanzania. The analysis was based on data obtained from 281 projects implemented by TANROADS in 11 regions from 2015 to 2025. Fourteen uncertainty factors were measured on a five-point Likert scale and screened using Principal Component Analysis (PCA), Variance Inflation Factor (VIF) and probability threshold (p2values of 0.9308, 0.9734 and 0.9233 for BV, SA, and NE respectively. The integrated framework of Likert scale uncertainty quantification, statistical pre-processing and ANN modelling showed an accurate KPI forecasting. The proposed approach provides both theoretical understanding of the effects of uncertainty factors on project performance and practical decision-support tools to mitigate cost overruns, schedule delays and quality deviations.
Recommended Citation
Mativila, H., Kafuku, J. M., & Kundi, B. A. (2026). Artificial Neural Network-Based Modelling and Prediction of Key Performance Indicators in Road Construction Projects. Tanzania Journal of Engineering and Technology (TJET), 45(2), 181-197. https://doi.org/10.65085/2619-8789.1284
Publisher Name
University of Dar es Salaam
Included in
Operations Research, Systems Engineering and Industrial Engineering Commons, Systems Science Commons