ACEVM applies machine learning to early-stage project-control forecasting

Enterprise analyticsExplainable AIResearch

Published: 25 July 2026. Haytham Elmousalami and Abdulaziz Alotaibi introduce Automated Conceptual Earned Value Management, trained and validated on 1,277 completed field-canal projects. Sixty optimised model instances predict early budget, duration, cost variance, schedule variance and project risk, with explainability and computational emissions included in model assessment.

Why it matters: Traditional earned-value management has little evidence before execution; the study moves forecasting into project conception. Impact: It offers a decision-support pattern for portfolio governance, but applicability beyond the source project class requires external validation. Peer-reviewed paper