Abstract Value Management (VM) of construction projects is beset by inherent pitfalls of expertise-dependence, fixed processes, and segregation from data-rich environments. The following paper presents and evaluates an artificial intelligence-facilitated Value Management System (AIVMS) that incorporates predictive analytics, Multi-Criteria Decision-Making (MCDM), and Explainable AI (XAI) to facilitate open, fact-based stakeholder-centric decisions throughout project life cycles. It was designed using the Design Science Research approach on systematic literature review of 127 peer-reviewed papers and was validated with three-round Delphi study with 24 construction professionals. The AIVMS system is six-layered and consists of: intelligent value driver identification, predictive analytics engine, dynamic MCDM engine, integration and optimization core, explainable AI interface, and adaptive learning system. Empirical validation through three real-world project case studies revealed significant improvements: 23% increase in decision-making consistency, 31% reduction in value engineering cycle time, and 89% improvement in stakeholder satisfaction with transparency of decisions. The framework achieved 91.2% precision for forecasting a variety of performance measures and enabled the identification of €2.8 M average cost optimization potential. This research is the first empirically-validated integration of AI, MCDM, and XAI for construction value management that integrates machine-based intelligence with man-centric transparency requirements and provides real-world implementation avenues for existing BIM and project management systems.
Mlybari et al. (Mon,) studied this question.