Methods capable of predicting the energy use and CO2 emissions of off-highway trucks, especially in the initial planning phase, are rare. This study proposed an artificial neural networks (ANN) model to assess such energy use and CO2 emissions for each unit volume of hauled materials associated with each hauling distance. Data from discrete event simulations (DES), an off-highway truck database, and different site conditions were simultaneously analyzed to train and test the proposed ANN model. Six independent quantities (i.e., truck utilization rate, haul distance, loading time, swelling factor, truck capacity, and grade horsepower) were used as the input parameters for each model. The developed model is an efficient tool capable of assessing the energy use and CO2 emissions of off-highway trucks in the initial planning stage. The results revealed that the grade horsepower and haul distances yield a significant increase in the environmental impact of the trucks. In addition, the results demonstrated that, for a given set of project conditions, the environmental impact of trucks can reduced by improving their utilization rate and reducing the loading time.
