Probabilistic Grading and Classification System for End-of-Life Building Components Toward Circular Economy Loop

2025-04-09

AI Summary

This study addresses the challenges of reuse decision-making and low value recovery rates for modular building components within circular economy frameworks. Methodologically, it proposes a data-driven reusability assessment framework featuring a five-level probabilistic classification system that integrates Bayesian modeling, scenario-driven thresholding, and dynamic feature weighting, coupled with interpretable decision trees and Sankey diagram-based traceability visualization to enable coordinated determination of reuse, upcycling, and downcycling pathways. Its key contribution lies in establishing, for the first time, an empirically grounded, multi-level probabilistic assessment paradigm that simultaneously satisfies engineering compliance and sustainability objectives. Validated on precast prestressed concrete wall panels, the framework significantly reduces construction material waste, enhances value recovery efficiency, and improves transparency and robustness in end-of-life (EoL) management decisions.

Technology Category

• Reasoning under Uncertainty: Relational Probabilistic Models
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Economics, Online Markets and Human Computation: Sustainability of Web economics
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

The longevity and viability of construction components in a circular economy demand a robust, data-informed framework for reuse decision-making. This paper introduces a multi-level grading and classification system that combines Bayesian probabilistic modeling with scenario-based performance thresholds to assess the reusability of end-of-life modular components. By grading components across a five-tier scale, the system supports strategic decisions for reuse, up-use, or down-use, ensuring alignment with engineering standards and sustainability objectives. The model's development is grounded in empirical data from precast concrete wall panels, and its explainability is enhanced through decision tree logic and Sankey visualizations that trace the influence of contextual scenarios on classification outcomes. MGCS addresses the environmental, economic, and operational challenges of EoL management--reducing material waste, optimizing value recovery, and improving workflow efficiency. Through dynamic feature weighting and transparent reasoning, the system offers a practical yet rigorous pathway to embed circular thinking into construction industry practices.

Problem

• Develops a grading system for reusing building components sustainably
• Uses Bayesian modeling to assess end-of-life component reusability
• Reduces waste and improves efficiency in construction recycling

Innovation

• Bayesian probabilistic modeling for reusability assessment
• Five-tier grading scale for component reuse decisions
• Decision tree logic with Sankey visualization