Life Cycle Modeling and Multi-Objective Optimization Framework for Low-Carbon Sustainable Construction Materials using Machine Learning: A Comprehensive Review

Life Cycle Modeling and Multi-Objective Optimization Framework for Low-Carbon Sustainable Construction Materials using Machine Learning: A Comprehensive Review

Authors

  • Gopinath Nallagarla, Dr.P.Sundar Kumar, Dr.Jerripotu Gopal Rao

Keywords:

Sustainable construction materials; Low-carbon concrete; Life cycle assessment; Multi-objective optimization; Explainable artificial intelligence; Geopolymers; Machine learning; SHAP; NSGA-II; Embodied carbon.

Abstract

The construction sector accounts for nearly 40% of global CO₂ emissions and approximately 50% of total resource consumption worldwide, underscoring an urgent imperative to transition toward low-carbon, sustainable material systems. This comprehensive review synthesizes over six decades of research progress across five interconnected domains that collectively define the frontier of modern sustainable construction science: (i) sustainable construction materials, including supplementary cementitious materials (SCMs), geopolymers, recycled aggregates, and bio-based composites; (ii) low-carbon strategies and decarbonization pathways encompassing alternative binders, carbon capture and utilization (CCU), and CO₂-cured concretes; (iii) life cycle analysis (LCA)-based modelling frameworks covering cradle-to-gate, cradle-to-grave, and dynamic LCA methodologies; (iv) multi-objective optimization (MOO) frameworks employing evolutionary algorithms, metaheuristic techniques, and Pareto-front analysis for concurrent minimization of cost, embodied carbon, and structural performance trade-offs; and (v) explainable artificial intelligence (XAI) methods—principally SHAP (SHapley Additive exPlanations), LIME (Local Interpretable Model-agnostic Explanations), and gradient-based attribution techniques—applied to materials informatics. Drawing on more than 65 peer-reviewed publications from leading journals, this review critically evaluates methodological advances, identifies persistent research gaps, and delineates a roadmap for integrating XAI-driven surrogate models with LCA and MOO into a unified decision-support framework. Key findings highlight that while each sub-domain has matured considerably, their holistic integration remains nascent, and a rigorous, transparent AI-LCA-optimization pipeline represents the most promising direction for accelerating the design and deployment of low-carbon construction materials at scale.

Downloads

Published

2026-09-27

Issue

Section

Articles

Citation Check

Loading...