Fuzzy Logic-Based Method for Forecasting Project Final Cost

Forecasting the final cost of construction projects during the construction phases is challenging, particularly for long-duration projects, due to the rise in uncertainties associated with future cost performance index values after report dates and the impact of many factors on cost performance. Current practices, along with existing methods and models, often assume that the cost performance index (CPI) achieved at the report date will continue as is for the remaining work and they fail to assess the risk of cost overruns. This assumption may not be true, as in many cases, the cost performance may change due to the impact of many factors. Thus, this paper aims to introduce a new multi-method fuzzy-based forecasting model to forecast the project’s final cost and to circumvent the limitations and setbacks of the used models and methods. It integrates earned value concepts with expert judgment using fuzzy logic to address the limitations of current practice in forecasting a project’s final cost during construction. Factors influencing future cost performance were identified through conducting in-depth literature reviews and expert interviews. The identified factors were then weighed using the relative importance index and pairwise comparisons. The top five most important factors are labor productivity, rework percentage, manpower availability, subcontractor management, and project plan controllability. These factors were then utilized to define the uncertainty associated with the future cost performance according to a project management team’s experience, using fuzzy numbers for forecasting the project’s final cost. Validation with real construction projects showed that the model provides more accurate predictions of completion costs through straightforward calculations. The model has been implemented in a computer application that is integrated with commercial scheduling software (P6 Professional).