Predicting the effects of congestion on labour productivity in construction projects using agent-based modelling

Abstract

Increasing the volume of labour is an approach taken by contractors to accelerate the workflow in construction projects. However, this will lead to congestion as a major contributing factor that can affect labour productivity and significantly impact the construction process. Various studies have analyzed the effects of congestion on labour productivity; however, they fail to quantify the impacts of congestion on labour productivity. Therefore, this paper applies the Agent-Based Modelling (ABM) as an effective tool for predicting the effects of congestion on labour productivity in construction projects. A Java-based application is developed as an implementation of the proposed model. Based on a real case study, the paper discusses the potential benefits of the model, e.g., accurately predicting labour productivity, providing opportunity for workspace allocation, and simulating project reality in a virtual model. The simulation results demonstrate a 30% decrease in productivity due to congestion. The results confirm that the impacts of congestion on labour productivity at a construction site can be predicted as an emergent property from the interaction between workers, tasks, and their environment, i.e., a construction site, as well as the dynamics of varying productivity.