AN AGENT-BASED APPROACH FOR MODELING THE EFFECT OF LEARNING CURVE ON LABOR PRODUCTIVITY BRIDGES

Abstract

The labor-intensive nature of construction projects requires proper management and efficient utilization of labor resources. Improvement of labor productivity can enhance project performance and thereby lead to substantial time and cost savings. Several studies focused on identifying the effect of different factors on labor productivity, whereby the learning curve factor proved of paramount importance. Although previous research efforts developed models to represent the learning curve effect using traditional simulation approaches such as System Dynamics (SD) and Discrete Event Simulation (DES), none of these studies used Agent-Based Modeling (ABM) techniques. This study takes the initial steps and presents work targeted at analyzing the effect of learning on labor productivity using ABM. Based on ABM, a construction site can be modeled as an active environment in which agents interact with each other and their surroundings thereby creating an adaptive environment open for learning and improvement. The solution to the problem is described in details using a simulation model developed in AnyLogic 7 (Educational Version). The components of the proposed model have been created and preliminary results highlighted the potential of using the agent-based modeling paradigm to simulate the effect of learning on labor productivity in the construction industry