A fuzzy expert system for soil characterization

Area of Science:

• Environmental Science
• Computer Science
• Soil Science

Background:

• Soil contamination poses significant risks to human health and ecosystems, necessitating robust risk characterization methods.
• Artificial Intelligence (AI) and Decision Support Systems (DSSs) offer advanced tools for managing contaminated sites.
• Traditional methods may lack the adaptability to handle the complexities and uncertainties in soil risk assessment.

Purpose of the Study:

• To develop and evaluate an AI-based Decision Support System (DSS) for characterizing contaminated soils.
• To classify soils based on associated human and environmental risks using an expert system approach.
• To provide an interpretable and scalable tool for contaminated soil management.

Main Methods:

• Development of a Soil Risk Characterization Decision Support System (SRC-DSS) using knowledge engineering.
• Integration of decision trees and fuzzy logic for enhanced data handling and interpretation.
• Utilizing 26 parameters categorized into source attributes, transfer vector attributes, and local properties.
• Evaluation of the SRC-DSS with sixteen diverse case studies.

Main Results:

• The SRC-DSS successfully classified and characterized contaminated soils according to identified risks.
• The system demonstrated excellent performance in risk assessment compared to other techniques.
• The tool proved effective in managing large-scale applications and handling data variability.

Conclusions:

• The developed SRC-DSS is a highly effective tool for classifying and characterizing contaminated soils.
• AI, particularly fuzzy logic and decision trees, enhances the accuracy and interpretability of soil risk assessments.
• The SRC-DSS provides a valuable framework for informed decision-making in contaminated soil management.