Predicting labor costs in construction projects using agent-based modeling and simulation

Abstract

In most construction projects, human resources account for 30 to 50 percent of all project costs, thus it is an important cost element in construction projects. However, the complexities and uncertainty of work ow have made it di cult to predict human resource costs. The most e ective and economic management approach is computer simulation which has a better ability to understand and identify the dynamics and statics of a case research system. One of the steps of simulation is modeling, and choosing the best modeling approach for prediction is very important. Agent-based modeling, which is referred to by some as a third way of doing science", was chosen to predict labor cost impacted by many factors in the work flow. In our research, the increased labor cost, based on factors a ecting work ow, is 16 percent on average, which is predicted more accurately by the simulated model than by planned cost.