25th ANNUAL WHEEL RAIL INTERACTION CONFERENCE

The Educational Railroading Conference Leader Since 1994

Beenvision Abstract:

Beenavision

Automated vision-based wayside inspection system - wheel tread & surface

Automated condition-based monitoring of train components has received considerable attention over the past two decades. As the wheelset is the most maintenance intensive component of the train, availability of a complete and automated inspection of the wheel at track speed will have a significantly positive impact on the maintenance practices of all train operators. In this Infozone, we present an entirely new approach for wayside wheel condition monitoring based on “machine vision technology.” This new approach is designed to measure and inspect wheel surfaces to a submillimeter resolution at mainline speed. This system scans the wheel to detect wheel surface defects such as shelling, spalling, grooves, broken and missing parts, wheel flats, out of roundness, providing a holistic view of the wheel condition. To thoroughly inspect the wheel, the system must be able to provide data from all around the wheel for a full 360 degrees. This means that the system must be capable of both imaging the wheel surface and the plate, and provide detailed 3D wheel surface data. This creates an enormous amount of data that needs to be processed to detect a variety of wheel defects.


Treadview System
Figure 1: Treadview System
Sample image of a broken wheel
Figure 2: Sample image of a broken wheel
3D wheel surface data
Figure 3: 3D wheel surface data