Risk-indexed artificial neural network for predicting duration and cost of irrigation canal-lining projects using...

Area of Science:

• Civil Engineering
• Project Management
• Artificial Intelligence

Background:

• Irrigation canal lining projects frequently face delays and budget overruns due to inherent uncertainties.
• Accurate estimation of project duration and cost is critical for successful infrastructure development.

Purpose of the Study:

• To develop and validate a risk-driven predictive model for estimating project duration and cost in irrigation canal lining projects.
• To create a practical decision-support tool for engineers and planners using an integrated framework.

Main Methods:

• Reduced 93 risk factors to 20 using Analytic Hierarchy Process-Revised Importance Index (AHP-RII).
• Trained a multi-layer perceptron artificial neural network (ANN) on 5000 simulated scenarios.
• Validated the model using leave-one-project-out cross-validation on eight real-world projects.

Main Results:

• The ANN model achieved high accuracy, with a testing R-squared of 0.82.
• Average prediction errors were within 0.87 months for time and EGP 102,500 for cost.
• The model was deployed as a user-friendly Python-based desktop application.

Conclusions:

• The integrated ANN-based framework effectively combines expert risk assessment with machine learning for improved project forecasting.
• The developed model provides a practical and accurate decision-support tool for early-stage infrastructure project planning.
• This approach enhances the reliability of time and cost estimations in complex engineering projects.