Computer vision can be used to classify sun kinks at speed from a train-mounted sensor.
The GPS-triangulated sun kink data can then be used to dispatch repair teams to the appropriate location preventing derailments in the process.
This has been studied academically and commercially in the US for a few years, at least. I'm not aware of a commercial deployment, but it may be in progress given the published R&D.
The challenge is ensuring you have sufficiently frequent passes on hot days, and therefore sufficient fitted trains, right?
That paper seems to be notable because it’s doing it from vision alone (and no LIDAR), whereas similar things (including gauging generally) have been used commercially for a while; see eg http://www.rail-vision.co.uk/
The GPS-triangulated sun kink data can then be used to dispatch repair teams to the appropriate location preventing derailments in the process.
This has been studied academically and commercially in the US for a few years, at least. I'm not aware of a commercial deployment, but it may be in progress given the published R&D.
For instance:
https://arxiv.org/pdf/1802.01286.pdf
From the paper: "Sun kink classifier could classify professionally simulated sun kink videos with a precision of 97.5%."