Saferail.tech

Infrared Detection of Animals
Example of UV on metal
UV image after processing and crack detection

Saferail.tech - Railway Obstacle Detection Device

Inspiration

In the recent decade, there have been numerous railway accidents involving train derailments. In 2016 alone, a shocking 537 train derailments had been reported to the Transportation Safety Board of Canada.[1] The majority of these incidents are a result of broken rails (15.3%), morphed track geometry (7.3%), and other rail related defects such as obstacles (27.2%). Therefore, 269 train accidents in 2016 in Canada involve the poor conditions of railroads.

** Objective of this project & Product details. ** The objective of this project is to design a safety device for any train, that is able to detect railroad defects and obstacles. This device analyzes the railroad ahead of the train and warns the locomotive crew of potential hazards. This device would be used on railroads with less frequent traffic since they are the most susceptible to poor maintenance and accidents.

What it does

The product consists of two robotic devices that roll along both rails in sync, similar to the wheels of a train. Both devices will be connected with a metal wire so that any obstacles in between the rails can be detected as well. The devices are battery powered and are aimed to last a minimum of 250km. The batteries would be easily replaceable, so that when the battery is low on charge, the engineers of the train can swap the batteries out. The device is powerful and slim enough to power through snow. An onboard micro-controller (raspberry-pi) analyzes the device’s surrounding environment, such as: • The condition of the rail (Such as any fractures or weak points) • The distance apart between the two rails • Any obstructions on or in between the rails • Any living organism within a certain distance of the tracks (using infrared technologies) • Improper position of railroad switches

How we built it

Designed circuit and hardware schematics using VHDL Wrote software for the analysis of environment in Java and Python Wrote presentation website in html, css and JavaScript (including libraries such as jQuery) Made demonstration in Autodesk Maya

Challenges we ran into

Getting training data for infrared and UV track sensors Data transferring Adapting parallax for mobile Losing wifi connection

Accomplishments that we're proud of

Integration and collaboration

What we learned

Image processing 5G technology requirements

What's next for Saferail.tech

-Design a platform where the engineers of the train can communicate to the railway obstacle detection device -Create a prototype -Train the UV and Infrared detectors on real training data -Propose idea to CN Rail, Amtrack, and other major railway corporations.

Sources

Government of Canada, Transportation Safety Board of Canada. “Railway Occurrences 2016 - Statistical Summary - Transportation Safety Board of Canada.” Government of Canada, Transportation Safety Board of Canada, 7 Dec. 2017, www.bst-tsb.gc.ca/eng/ stats/rail/2016/sser-ssro-2016.asp