Intermodal trains are typically the fastest trains operated by North American freight railroads. It is thus ironic that these trains tend to have the poorest aerodynamic characteristics. Because of constraints imposed by equipment design and diversity, there are often large gaps between intermodal loads and these trains incur greater aerodynamic penalties and increased fuel consumption compared to other trains. We conducted train energy analyses of the most common intermodal train configurations operated in North America. It was found that matching intermodal loads with cars of appropriate length reduces the gap length thereby improving airflow. Properly matching cars with loads also avoids use of cars that are longer and thus heavier than necessary. For double stack containers on well cars, train resistance may be reduced by as much as 9% and fuel savings by 0.52 gallon per mile per train. Proper loading of intermodal trains is therefore important to improving energy efficiency. We have developed a wayside machine vision system that automatically scans passing trains and assesses the aerodynamic efficiency of the loading pattern. Machine vision algorithms are used to analyze these images and detect and measure gaps between loads and develop a quantitative index of the loading efficiency of the train. Integration of this metric that we call "slot efficiency" can provide intermodal terminal mangers feedback on loading performance for trains and be integrated into the software support systems used for loading assignment.