to AI-Augmented Earned Value Forecasting:
Practical Lessons from Complex Projects
ADVISORY ARTICLE
By Tauseef Naz Arshad
Reading, United Kingdom
Abstract
Earned Value Management (EVM) and the Critical Path Method (CPM) remain central to project control practice. In complex engineering, procurement, and construction (EPC) and infrastructure projects, however, practitioners often observe that EVM-based forecasts remain optimistic well into execution, even as delivery risk continues to increase. This article argues that a key contributor to this behavior is critical path volatility, which is frequently present but rarely treated explicitly in forecasting logic. Based on practical industry experience, the article explains how repeated changes in critical path structure undermine core EVM assumptions. It then outlines how AI-augmented forecasting approaches can improve predictive reliability by incorporating schedule volatility signals alongside traditional performance metrics. The emphasis is practical, with the objective of improving forecast credibility, decision timing, and management confidence in live project control environments.
- Introduction
Earned Value Management (EVM) and the Critical Path Method (CPM) have long formed the foundation of project performance monitoring. In relatively stable project environments, these techniques can provide reasonable visibility into cost and schedule outcomes. In large EPC and infrastructure programs, however, project control teams frequently encounter a different reality.
It is common for EVM indicators to remain stable for extended periods, creating an impression that delivery is broadly under control. Later in execution, forecasts can deteriorate rapidly, often at a point where schedule flexibility has already reduced, and recovery options are limited. This pattern is not always the result of optimistic reporting or weak controls. In many cases, it reflects a deeper structural issue within the schedule itself.
One of the most significant contributing factors is critical path volatility. As execution progresses, the activities driving project completion can change repeatedly due to resequencing, late engineering information, procurement delays, or recovery actions. Traditional EVM forecasting methods, however, assume that schedule structure remains broadly stable and that past performance can be projected forward with limited adjustment. When these assumptions no longer hold, forecast accuracy degrades in a systematic way.
This article draws on practitioner experience to explain why critical path volatility undermines EVM forecasts and how AI-augmented approaches can help address this limitation in complex project environments.
More…
To read entire article, click here
How to cite this article: Arshad, T. N. (2026). From Critical Path Volatility to AI-Augmented Earned Value Forecasting: Practical Lessons from Complex Projects, PM World Journal, Vol. XV, Issue III, March. Available online at https://pmworldjournal.com/wp-content/uploads/2026/03/pmwj162-Mar2026-Arshad-From-Critical-Path-Volatility-to-AI-Augmented-EVM.pdf
About the Author
Tauseef Naz Arshad
Reading- United Kingdom
Tauseef Naz Arshad (PMP, RMP, PMI-SP, ACP, PgMP, PfMP) is a senior project planning and project controls professional with over twenty years of experience delivering large-scale Engineering, Procurement, and Construction (EPC) programs across complex industrial and infrastructure environments. His work focuses on the practical integration of advanced analytics, artificial intelligence, and digital technologies into project scheduling, forecasting, and executive decision support. He has led planning and project controls functions on major EPC projects and continues to explore data-driven approaches that bridge established industry practices with emerging digital capabilities. Mr. Arshad can be contacted at tauseefnaz@gmail.com
