Page 12 - Spring 2024

RAIL UPDATE

McLaren helps maintain Railway with 5G The solution, engineered by Tokyo Artisan
Connectivity Intelligence and powered by AMD, uses rapid
image processing coupled with AI to identify and
McLaren Applied is helping Network Rail to
examine issues such as loose bolts and other track
maintain the railway with the installation of
anomalies. Central to this is a vision computing
its F1-derived Fleet Connect software and 5G
box mounted on a cart which conducts track
Edge Active Antenna hardware on 14 of its
inspections at 12 mph. A high-speed camera is
infrastructure monitoring trains.
incorporated into the unit along with an AMD Kria
K26 System-on-Module that provides the AI-
This first implementation of Active Antenna
powered data and image analysis both before
technology in the UK brings a new, simplified
and after capture. The Kria K26 is a compact,
approach to on-train connectivity infrastructure
integrated platform that incorporates a custom
and an advanced capability for data offloading in
AMD Zynq UltraScale+ MPSoC, DDR memory,
the move.
non-volatile storage, a security module and an
The UK’s infrastructure monitoring trains travel aluminium thermal heat spreader.
across the country checking the condition of
The AMD Kria SoM’s programmability, durability
bridges, tunnels, viaducts, signals, level crossings
and embedded intelligence has allowed it to
and around 20,000 miles of track. These trains
adapt to the diverse and evolving landscapes,
report back using live data streams including
conditions and demands of JR Kyushu’s extensive
sensor readings on the track, foliage and network
territory and passenger base with over 1,455
monitoring, and the remote and automated
km of track. The dynamic natural environment
monitoring helps to target maintenance work and
of railway tracks necessitates the capacity for
maintain safety and reliability for rail services.
ongoing updates to the Kria SoM in order to ensure
The patented Fleet Connect software splits the longevity and relevance of the technological
data across networks in real-time and then investment.
reassembles it in the cloud, reducing blackspots
Chetan Khona, Senior Director of Industrial, Vision,
and providing a consistent connectivity stream for
Healthcare and Sciences Markets at AMD, states:
any moving vehicle in any environment.
“ AMD Kria SoMs are accelerating innovation at
Since the introduction of this new technology
the Edge and simplifying solution development
Network Rail has registered upload data
through deployment. JR Kyushu is a perfect
improvements of up to a thousand times faster
example of how the endless programmability
than previous systems with speeds of up to 70
of Kria SoMs, combined with Edge AI computing,
MB/s per train.
can automate processes and dramatically
Reference: Fowle, H., Electronic Specifier, 16th. January enhance operational efficiency across a variety
of applications, from machine vision to industrial
robotics and AI/ML computing. “
AI revolutionises Track Inspection for Bullet Train
Reference: West, P., ‘AMD powers AI Track Inspection Solution for
AMD has joined with the JR Kyushu Railway Japanese Bullet Train’, Electronic Specifier, 14th. February
Company (JR Kyushu), a prominent operator of
the Japanese Bullet Train to bring AI to its track
inspection.
12
Subscribe to Electron here https://membermojo.co.uk/ie