A Data-Driven Framework for Forecasting Project Performance

In the dynamic field of construction project management, accurately forecasting project performance remains a significant challenge. Budget overruns and schedule delays are common issues, and despite the abundance of available data, existing forecasting practices struggle to effectively leverage this information. This study addresses gaps in the current literature by proposing a data-driven approach to enhance Earned Value Management (EVM). EVM is a method for forecasting both project cost and performance, incorporating key metrics such as the estimate at completion (EAC). The research introduces a comprehensive framework tailored for the execution phase, utilizing various data sources such as historical data, project data from schedule updates, and expert opinions. The proposed research unfolds in two phases: Phase 1 involves collecting, synthesizing, and preparing data for a data-driven model, while Phase 2 focuses on developing a data-driven approach for forecasting project schedules, leveraging Bayesian statistics and discrete event simulation. Additionally, it integrates the forecasted schedule’s impact on project cost prediction using a rule-based fuzzy logic system. The overall methodology aims to enhance project forecasting in EVM, providing valuable insights for effective construction project management. This collaboration provides SMA Consulting with the opportunity to leverage cutting-edge solutions, enhancing competitiveness and reinforcing their commitment to excellence. The adoption of superior tools resulting from this research aligns seamlessly with SMA Consulting’s commitment to delivering excellence in project management and controls, positioning them as an industry leader.

Faculty Supervisor:

Yasser Mohamed

Student:

Partner:

SMA Consulting

Discipline:

Engineering

Sector:

Construction and infrastructure

University:

University of Alberta

Program: