Adaptive Forecasting in Earthmoving Operation Using DES and Site Captured Data

Abstract

Most forecasting techniques for repetitive construction operations use deterministic approaches or stochastic approaches that are based on historical data of similar past projects. Utilizing such data impacts the accuracy of simulation results as it does not account for the unique characteristics of the construction operation being considered. This paper presents an automated tool for earthmoving operations that performs adaptive Discrete Event Simulation (DES) through utilizing actual captured data. Adaptation is curried out utilizing actual captured data from earthmoving operations to forecast productivity and consequently time and cost of these operations. The data was captured by GPS mounted on earthmoving hauling trucks then transferred wirelessly to a central server. A specially designed relational database was developed to store the captured data in addition to the characteristics of the earthmoving operation. The captured data implicitly accounts for factors that give rise to uncertainty such as weather. The durations captured for loading, hauling, dumping and returning activities are used to generate representative Probability Distribution Functions (PDF). DES is used to forecast fleet productivity and consequently the forecasted activity's duration and cost. In this research “StroboScope” DES engine was integrated with the relational database and Oracle Crystal Ball in standalone software.