Level: Advanced
Venue: 2019 AACE International Conference & Expo, June 16-19, 2019, New Orleans, LA, USA
Abstract: The S-curve is the common tool used for depicting the project cumulative progress during execution. Numerous studies have been found to tackle the forecasting of the project-level cash flow in the preconstruction stages, using prediction techniques involving neural network and regression analyses as well as third-degree polynomial and sigmoid functions. The encountered models rely on an array of input variables, including the type of work and location, degree of project simplicity, team competence, curve slope and inflection point, and specific time-money milestones, among others. The objective of the presented work is concerned with the investigation of the applicability and prediction accuracy of the proposed planned progress estimation models from the perspectives of construction project owners and their appointed contract engineers. In this regard, data related to the earned value figures achieved from work progress actually made on a completed residential project were used for forecasting the contractor’s S-curve using the adopted models. The performed sensitivity analyses allowed the determination of those ranges of the input factors that satisfy an acceptable degree of prediction accuracy. The findings revealed a perfect fourth-degree polynomial fit to both the upper and lower curves forming the acceptable progress envelope deduced from the performed sensitivity analysis.
Venue: 2019 AACE International Conference & Expo, June 16-19, 2019, New Orleans, LA, USA
Abstract: The S-curve is the common tool used for depicting the project cumulative progress during execution. Numerous studies have been found to tackle the forecasting of the project-level cash flow in the preconstruction stages, using prediction techniques involving neural network and regression analyses as well as third-degree polynomial and sigmoid functions. The encountered models rely on an array of input variables, including the type of work and location, degree of project simplicity, team competence, curve slope and inflection point, and specific time-money milestones, among others. The objective of the presented work is concerned with the investigation of the applicability and prediction accuracy of the proposed planned progress estimation models from the perspectives of construction project owners and their appointed contract engineers. In this regard, data related to the earned value figures achieved from work progress actually made on a completed residential project were used for forecasting the contractor’s S-curve using the adopted models. The performed sensitivity analyses allowed the determination of those ranges of the input factors that satisfy an acceptable degree of prediction accuracy. The findings revealed a perfect fourth-degree polynomial fit to both the upper and lower curves forming the acceptable progress envelope deduced from the performed sensitivity analysis.