Bridging Theory and Prediction: A Hybrid <scp>SEM</scp> and Machine Learning Approach to Optimize Lean Construction for Megaproject Sustainability in China

ABSTRACT Construction megaprojects, large‐scale, complex, and capital‐intensive, are particularly prone to inefficiencies, cost overruns, delays, and environmental degradation due to fragmented workflows, stakeholder misalignment, and resource intensity. Lean Construction Practices (LCPs), with their focus on waste elimination, value stream optimization, and collaborative planning, offer a targeted response to these megaproject‐specific challenges, directly supporting sustainability goals. However, empirical evidence on the systematic integration of LCPs into sustainable megaproject delivery, especially in emerging economies, remains sparse. This study addresses this gap by (1) quantifying the causal impact of LCPs on Overall Sustainable Success (OSS), defined as a tripartite construct encompassing environmental resilience (e.g., waste/energy reduction, pollution control), social inclusivity (e.g., worker safety, collaboration, equity), and economic efficiency (e.g., cost control, rework minimization, productivity gains), using PLS‐SEM; and (2) identifying the most predictive LCPs for OSS using explainable Machine Learning (ML), with a focus on China, the world's largest megaproject market. Using survey data from 379 randomly sampled professionals engaged in megaprojects across Mainland China and Hong Kong, results confirm a strong LCP‐OSS relationship ( β = 0.748), with Gradient Boosting achieving 82% accuracy and 88% ROC‐AUC. Crucially, SHAP analysis is innovatively applied at both indicator and construct levels, enabling actionable prioritization of LCPs, a methodological advance for sustainability research. Top practices include Safety and Quality Assurance (22.2%), Customer Focus and Waste Elimination (20.8%), and Standardization and Process Transparency (18.8%). While contextually grounded in China, findings align with SDGs 9, 11, and 12, suggesting transferability to similar emerging economies. The framework provides policymakers and practitioners with evidence‐based levers to integrate sustainability into megaproject delivery, without compromising efficiency or equity.