AI-based defect detection software provider Ardenna has developed the capability to identify railway track defects using software on a UAV payload.
The technology is designed to ensure railway track integrity in real-time, including detection of any potential obstructions, rail breaks and track buckling.
Once deployed, the application will enable rail operators to initiate prompt measures to resolve railway track issues, significantly improving overall safety and services.
The process involves taking visual images of the railway tracks by a camera placed on the UAV. Images are then processed by dedicated software algorithms running on an NVIDIA Jetson TX2 embedded computing module.
The system generates instant alerts when anomalies are found.
Ardenna business development director David Patterson said: “We are excited to be the first to demonstrate the ability to embed an automated defect detection capability in a form factor suitable for drone payload integration.
“As we continuously enhance the capabilities of our Cloud and desktop solutions for the Transportation and Energy Sectors, we will naturally bring more of the core functionality to our embedded solution.
“These capabilities will only be enhanced as computing power and drone battery-life increases.”
The application can be used by train operators to carry out routine inspections, as well as assess the rail track conditions after natural disasters such as floods and storms.
In addition, it can deploy the UAV payload before transporting any hazardous cargo.
This week, Ardenna demonstrated the application at Commercial UAV Expo in the Las Vegas, Nevada, US.