A Comparative Analysis of Work-Hour Forecasting Techniques at the Crew Level
Abstract
Several different methods can be used to forecast the productivity of labor-intensive construction activities. The results from three widely used method, percent complete, learning curve and standard productivity curve were used to test the practicality of a new forecasting method, the Factor Model, for a single masonry project. Forecasts were made at weekly intervals and then compared with the actual productivity at project completion. All of the methods produced divergent forecasts throughout the first third of the activity. However, the factor model produced a forecast within 4% of the final productivity after only 3% of the work had been completed. While this research did not show that the factor model produced more accurate forecasts than the other methods, it did show that the forecasts were equally accurate. Thus, it appears that the factor model is a plausible alternative to conventional forecasting techniques.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1989
- Accession Number
- ADA220706
Entities
People
- James M. Pace
Organizations
- Pennsylvania State University