This was the first interdisciplinary project, in collaboration with the Civil Engineering Railroad Program, for developing a machine vision system to perform railcar inspection. The railcar “truck” components detected for inspection where the wheel, the bearing endcap bolts, the spring nest, and the break shoe. | Presentation | Website | Publications |
Category: Automated Inspection
Machine Vision Inspection of Railcar Safety Appliances
Inspectors and carmen use safety appliances to mount and dismount railcars during yard operations. It is critical for their safety that these laters, handholds and sillsteps are in good condition. A machine vision system was designed and later implemented by two commercial vendors for operation on sevaral major railroads. | Presentation | Website | Publications […]
Multi-Spectal Machine Vision for Passenger Undercarriage Inspection
The use of both visible and infrared imaging was used to investigate the undercarriage of passenger trains isolate incipient failures and detection of foreign objects. The system has been tested on in-service passenger trains at the Amtrak Repair Facility in Chicago IL. | Presentation | Website | Publications |
A Machine Vision System for Railcar Underbody Structural Inspection
Heavy haul railcars must be visually inspected frequently to ensure the structural stability of the railcar underbody. The goal of this project was to determine a method for inspecting underneath moving railcars. A commercial system based on this work has been developed and is commercially available. | Presentation | Website | Publications |
Intermodal Load Monitoring Using a Wayside Machine Vision System
This project involved the monitoring of the loading of intermodal trains to provide an aerodynamic analysis of the entire train to estimate its fuel efficiency. Two wayside sites were developed and tested with in-service trains on a the BNSF railroad mainline in Sibley MO. | Presentation | Website | Publications |
Track Inspection Using Machine Vision
This project focused on automatically inspecting the condition of the track by identifying defective components, such as track fasteners and turnout components, using computer vision. Video was acquired by mounting cameras on a custom-build track vehicle. | Presentation | Website | Publications |