Recent derailments on the UK rail network have understandably brought renewed attention to track condition and railway resilience.
The exact causes of these incidents are still subject to investigation, and it would be wrong to speculate about their conclusions.
But they come at a time when another challenge facing Britain’s railway is becoming increasingly important: extreme heat and the resilience of the track infrastructure beneath our trains.
Network Rail reported in 2025 that hot weather had caused rail buckles resulting in more than 350,000 minutes of delays during the previous year. It has also committed £2.8 billion over five years to improving the network’s resilience to climate-related issues.
So what causes a railway to buckle, and could remote sensing help us better identify where that risk is greatest?
Heat is only part of the equation
Steel rails expand as they get hotter.
Network Rail states that rails exposed to direct sunlight can become as much as 20°C hotter than the surrounding air temperature. On continuously welded rail, those temperature changes generate substantial forces within the track. But heat alone does not determine whether the railway buckles.
As Network Rail itself explains:
“The mass of the sleepers and ballast are designed to contain these forces and prevent the track from buckling.”
That distinction is important, because temperature creates the force and the track structure has to resist it. The critical component providing that resistance is ballast.
Why ballast matters
Following the record temperatures experienced in 2022, Network Rail commissioned an Extreme Heat Task Force to examine how the railway could become more resilient to increasingly extreme temperatures.
Its subsequent Engineering Report provides an important insight into the causes of track buckling. The report states explicitly that:
“Heat alone, and the associated expansion of rail, is not the only factor that causes track to buckle.”
Among the principal factors and triggers it identifies are:
- ballast consolidation around sleepers and bearers;
- ballast condition;
- ballast profile and volumes, particularly between sleepers and at the ends and shoulders;
- drainage;
- sleeper condition; and
- rail age, type and stressing history.
In other words, understanding where the railway may be vulnerable isn’t simply a question of knowing how hot the rail is. We also need to understand how capable the track structure is of resisting the forces being generated.
Can we see some of that risk from the air?
At Network Mapping, we’re perhaps best known for our work with electrical utilities, where airborne LiDAR and high-resolution imagery are used to map and analyse thousands of kilometres of critical infrastructure, but our experience with large-scale linear infrastructure extends well beyond electricity. We have mapped the UK’s rail network using airborne LiDAR. That experience raises an interesting question:
Could airborne LiDAR be used not simply to map railway infrastructure, but to help identify areas where ballast profile and the surrounding track environment may contribute to increased susceptibility to instability?
Modern airborne LiDAR produces a highly accurate three-dimensional representation of an infrastructure corridor.
With sufficiently dense and accurate acquisition, analysis of that point cloud could potentially provide information about:
- ballast shoulder width and height;
- overall ballast profile;
- apparent areas of ballast deficiency;
- changes in ballast profile between surveys;
- track and embankment geometry;
- settlement or movement in surrounding earthworks; and
- drainage and vegetation conditions around the railway.
This would not replace engineering inspection or technologies designed to understand what is happening beneath the ballast, but it could provide another important layer of information, and potentially do so across very large areas, helping prioritise manual inspections.
Beyond the point cloud
LiDAR doesn’t need to work in isolation. Combining high-density airborne LiDAR with high-resolution 4-band imagery, RGB plus Near-Infrared (NIR), can provide additional information about the wider environment surrounding the track.
While LiDAR provides the precise 3D geometry needed to understand ballast profiles, track geometry and earthworks, multispectral imagery can provide another layer of insight into the condition of the railway corridor. Near-Infrared data can be used to derive vegetation indices such as NDVI, helping identify changes in vegetation health across embankments and cuttings. Changes or areas of stress can provide useful indicators for further investigation, particularly when considered alongside LiDAR-derived information on slope geometry, drainage and changes in the surrounding terrain.
High-resolution imagery also provides valuable visual context for anomalies identified within the point cloud, allowing engineers and asset managers to review locations remotely before determining whether further inspection is required.
The real opportunity therefore isn’t necessarily LiDAR alone. It’s the combination of 3D geometry, multispectral imagery and repeatable analytics to build a more complete picture of how the railway corridor is changing.
From temperature monitoring to susceptibility mapping
Network Rail already undertakes extensive preparation for periods of high temperature.
It uses weather forecasting, trackside temperature probes and remote monitoring, checks track stability ahead of summer, strengthens vulnerable areas and introduces speed restrictions when necessary.
Those measures are important because slower trains exert lower forces on the railway when the track is already under significant thermal stress, but there may be an opportunity to add another dimension to that approach.
Imagine combining:
Airborne LiDAR-derived information
- ballast profile
- track geometry
- earthworks
- drainage
- change over time
with:
Existing railway information
- rail temperature
- weather forecasts
- rail stressing records
- maintenance history
- known earthworks issues
Instead of simply asking:
“Where will the rail become hottest?”
we could begin asking:
“Where is the railway least able to resist the forces generated when it becomes hot?”
That moves the conversation from monitoring temperature towards mapping susceptibility.
Change may be as important as condition
There is another important advantage to remote sensing: repeatability. A single survey tells us what something looks like today; however, repeat surveys tell us what is changing.
Accurately aligned LiDAR datasets could potentially highlight sections where ballast profiles are changing, earthworks are moving or other characteristics of the railway corridor are developing differently from the surrounding network. This matters because infrastructure risk isn’t necessarily static.
Network Rail’s own climate resilience work acknowledges that track disturbance and maintenance activity, particularly activity affecting ballast, can temporarily increase instability until the track has fully consolidated.
That suggests that understanding change over time could be just as valuable as identifying whether a particular measurement falls outside a fixed threshold.
Rather than inspecting every kilometre equally, remote sensing could potentially help engineers identify the locations where something is changing and prioritise further investigation.
Why airborne?
Ground-based and train-mounted inspection technologies will continue to be essential to railway maintenance and airborne LiDAR shouldn’t be viewed as a replacement for them, however an aircraft can capture very large sections of railway rapidly and consistently without requiring track possession or placing equipment and personnel on the operational railway interrupting freight or passenger services
Furthermore the same acquisition captures considerably more than the ballast:
- Track geometry
- Embankments
- Cuttings
- Drainage environments
- Vegetation
- Structures and the wider railway corridor can all form part of the same three-dimensional dataset
That creates the possibility of moving beyond individual asset inspection towards a more holistic understanding of the environment in which those assets operate.
From mapping infrastructure to understanding risk
Network Rail’s own Extreme Heat Task Force makes the important point that heat alone does not cause track buckling.
It is the interaction between temperature, rail stress, ballast, sleepers, drainage, earthworks, maintenance history and other factors that determines the resilience of the track system.
That is fundamentally a geospatial problem.
We already know that airborne LiDAR can map railway infrastructure at network scale…because we’ve done it.
The next question is whether advances in point-cloud analytics, automated feature extraction and AI powered change detection can turn that mapping data into something more valuable:
a network-scale view of where track instability may be more likely to develop.
The investigations into the recent derailments should be allowed to reach their conclusions, but irrespective of those findings, the wider engineering challenge remains.
As Britain’s railway adapts to increasingly extreme temperatures, could combining airborne LiDAR and AI provide another tool to help infrastructure managers understand not simply where the network is getting hot, but where it may be most vulnerable when it does? If you’re working on track resilience or climate adaptation in the rail sector, we’d be glad to hear your perspective – get in touch to continue the conversation.