Aerodynamic Analysis of Intermodal Freight Trains Using Machine Vision

Intermodal freight transportation is one of the largest sources of revenue for North American freight railroads and has experienced considerable growth over the past few decades. However, intermodal trains use rolling stock that generates significantly greater aerodynamic drag compared to other types of freight trains. The increased drag results in greater fuel consumption, which increases annual operating expenditures. There are opportunities to improve intermodal train aerodynamics by strategically placing the intermodal loads within the train’s consist. A machine vision system is being developed to automatically monitor and analyze an intermodal train’s aerodynamic efficiency based on the container/trailer loading pattern and the gap lengths between them. This system’s main components are train detection sensors, a digital camera, video acquisition software, machine vision and analysis software, and a communication network. An automated system coordinates these components for the video capture of in-service trains. The machine vision algorithms separate the train from the background in the video and assemble a panoramic image of the entire train. Using this panorama, the containers, trailers, and gaps between the loads are identified and measured. Following the successful development of a prototype system at an intermodal terminal in Illinois, USA, a fully automated machine vision system is being developed along the BNSF Railway’s intermodal corridor from Chicago to Los Angeles in the USA. The outputs of this system include load pattern monitoring, gap length information, and aerodynamic scoring, which are used to evaluate the loading efficiency of each train. In addition to the machine vision system, research is being conducted to determine how intermodal terminal managers can improve their decision making so that intermodal train loading can be more energy efficient.