Comparison of LBMS schedule forecasts to actual progress

Abstract

Location-based Management System (LBMS) uses actual progress, resource, and productivity data to calculate schedule forecasts. The forecasting method has been developed by using empirical results from real construction projects and these forecasts are intended to be used as early warnings to highlight production problems and interference before they occur. It is expected that control actions are taken in response to adverse forecasts to prevent the forecasted problems from happening on site. The updated forecasting system based on earlier research has been used in several projects and now there is enough data to evaluate how well the forecasts reflect reality. In this research, forecasts from two hospital construction projects were evaluated for accuracy. Tasks were selected for analysis based on their manhour count, availability of accurate resource data and full completion of the operation. Forecasts were reviewed at four dates during performance of each task. The forecast on each date was compared to actual progress on these four dates. Results show that certain deviations such as working out of sequence, not finishing locations or working in several locations at the same time severely impact the forecasting ability of LBMS and make the project schedules unpredictable. Forecasting is accurate in the short term if there are no special circumstances and the work proceeds continuously.