Railroad Turnout Wear Diagnostics

Area of Science:

• Railway Engineering
• Mechanical Engineering
• Computer Vision

Background:

• Railway turnouts are critical infrastructure where approximately 90% of railway accidents occur.
• Understanding wear patterns and forces in turnouts is essential for preventing failures and ensuring safety.
• Existing diagnostic methods may not provide real-time, detailed analysis of turnout conditions.

Purpose of the Study:

• To present issues related to the technical condition of railway turnouts.
• To compare normal forces during vehicle traffic on straight tracks versus turnouts.
• To introduce an online diagnostic system using a vision system for detecting turnout wear.

Main Methods:

• Analysis of normal forces in wheel-rail contact for straight tracks and turnouts.
• Presentation of turnout wear processes at selected speeds.
• Development and implementation of image processing algorithms for wear detection using a vision system in MATLAB.
• Testing algorithms for generating wear areas from high-resolution images.

Main Results:

• Comparison of normal forces indicating increased stress on turnout components.
• Identification of turnout wear processes influenced by vehicle speed.
• Successful development of algorithms for online detection of wear and distances in turnout elements.
• Demonstration of the system's capability for real-time diagnosis of turnout conditions.

Conclusions:

• The proposed vision system offers an original, online method for diagnosing railway turnout elements.
• The system effectively detects wear and critical distances, contributing to improved railway safety.
• The developed algorithms are compatible with existing railway traffic control systems, enabling seamless integration.