Machine learning-based prediction of construction and demolition waste generation in developing countries: a case...

Area of Science:

• Environmental Science
• Civil Engineering
• Data Science

Background:

• Limited data on construction and demolition waste (C&DW) generation hinders effective management, especially in developing countries.
• Informed decision-making for waste management requires reliable forecasting of C&DW quantities.
• Construction and demolition activities generate significant waste streams that need strategic planning.

Purpose of the Study:

• To develop a reliable artificial intelligence (AI)-based model for forecasting monthly C&DW generation.
• To address the data scarcity challenge in C&DW waste management through predictive modeling.
• To provide a tool for policy and decision-making in Tehran, Iran's C&DW management.

Main Methods:

• Trained and evaluated multiple AI algorithms including multilayer perceptron neural network, radial basis function neural network, support vector machines, and adaptive neuro-fuzzy inference system (ANFIS).
• Utilized historical data for training prediction models to forecast C&DW generation.
• Assessed model performance using metrics such as R-squared (R²) and Root Mean Square Error (RMSE).

Main Results:

• All AI algorithms demonstrated high prediction performance for C&DW forecasting.
• The ANFIS model achieved the best performance, with R² = 0.96 and RMSE = 0.04209.
• ANFIS excelled due to its integration of fuzzy logic capabilities with neural networks for modeling subjective variables.

Conclusions:

• The developed AI-based prediction model, particularly ANFIS, offers a reliable method for forecasting C&DW generation.
• This predictive tool can significantly aid policymakers and stakeholders in making informed decisions for C&DW management.
• The study highlights the potential of AI in addressing data limitations for environmental management challenges.