Truck productivity is recognized as being a major cause of cost and time overruns in earthmoving operation. Traffic congestions in large construction sites might negatively affect truck productivity. This paper presents an intelligent agent-based model to improve truck productivity in construction sites with traffic congestion conditions. Reinforcement learning theory is adopted to train the agent. The development is carried out using MATLAB. A work example is provided to show the applicability and efficiency of the proposed model. The paper proves model's accuracy. This paper provides an artificial intelligence solution for truck productivity; therefore the paper contributes to automation in construction.
