Rockfall detection systems for railways monitor exposed cuttings, embankments, and tunnel portals for falling rock, then alert operators so trains can be slowed or stopped before impact. The critical distinction is between systems that monitor (collect data for later analysis) and systems that alarm (trigger action within seconds). For safety-critical railway operations, only real-time alarming protects trains in the moment a rockfall happens.
That distinction decides whether a system prevents an accident or merely documents one. This guide explains the problem, walks through the main detection technologies and where each is strong and weak, and sets out what to look for when specifying a system for an operational line.
The critical question is not whether to detect rockfall, but how fast and how reliably.
— Key principle for railway safety managers
Why rockfall is a structural risk to railways, not an occasional nuisance
Railways follow valleys, cut through hillsides, and exit tunnels directly beneath steep, fractured rock. These are precisely the locations where rockfall occurs. The operational exposure is severe for three reasons:
Reaction time is near zero. A train exiting a tunnel at 200 km/h covers more than 50 metres per second. A driver who sees an obstruction on the track has no meaningful time to react. Detection has to happen before the train arrives, not when the driver sees the rock.
Catch fences degrade silently. Energy-absorbing barriers are rated to fixed energy limits (for example 80 kJ or 100 kJ). Repeated loading from snow, debris, and minor impacts fatigues them. A barrier that performed at installation may not perform after a decade — and there is often no visual indication that its capacity has dropped.
The exposure is climate-driven and growing. Freeze-thaw cycles, intense rainfall, and slope saturation are increasing the frequency of spontaneous events. Operators are managing a rising baseline of risk across ageing earthworks.
Ignored, the consequences run from service delays and line closures to derailment and loss of life. The question for any operator with exposed alpine, coastal, or cutting-heavy corridors is not whether to detect rockfall, but how fast and how reliably.
Monitoring vs. alarming: the distinction that matters most
This is the single most important concept when specifying a railway rockfall system, and it is the one most often blurred in vendor literature.
Monitoring tells you a slope is deteriorating. Alarming tells you a rock is on the track right now. When you evaluate a system, ask: how many seconds from impact to actionable alarm?
Monitoring observes a slope over time. It builds a record of movement, vibration, or displacement that engineers analyse to understand long-term stability. It is valuable for surveying, trend analysis, and predicting where attention is needed.
Alarming detects an event as it happens and triggers an immediate response — an SMS to maintenance staff, a signal to a control centre, an automated line closure. Alarming creates action within seconds.
Monitoring tells you a slope is deteriorating. Alarming tells you a rock is on the track now. A safety-critical railway needs the second capability, and many systems marketed for “rockfall monitoring” do not provide it with the speed or reliability that operational safety requires. When you evaluate a system, ask one question first: how many seconds from impact to actionable alarm, and how is that figure proven in the field?
How many seconds from impact to actionable alarm — and how is that figure proven in the field?
— The question every procurement team should ask first
The main rockfall detection technologies, compared honestly
Each technology below has legitimate applications. The issue is matching the technology to the job — and continuous, safety-critical alarming on an operational line is a demanding job that exposes the limitations of several otherwise useful approaches.
| Technology | Real-time alarm | Weather-proof | Off-grid capable | Event certainty |
|---|---|---|---|---|
| Radar | ✗ Minutes delay | ✓ | ~ Partial | ~ Medium |
| DFO (Fibre) | ~ Variable | ✓ | ✗ | ✗ Low |
| LiDAR | ✗ | ✗ Rain/fog fail | ✗ Mains needed | ~ Medium |
| ImpactSentinel™ | ✓ 2–4 seconds | ✓ IP67 | ✓ Solar/battery | ✓ High |
Radar
Radar measures slope movement remotely and is genuinely useful for wide-area surveying and slow-movement analysis across a hillside.
For real-time railway alarming, two weaknesses matter. First, many radar deployments transmit over 4G/5G or satellite pathways, which can introduce delays of several minutes between event and alert — far too slow for a train. Second, a single hardware fault or alignment problem can take an entire monitored zone offline, with no redundancy. Radar earns its place in analysis. It is not a dependable last-line alarm.
Distributed fibre optic sensing (DFO)
DFO runs vibration sensing along kilometres of fibre cable, which is attractive for long rail corridors and gives excellent linear coverage.
The weakness is event certainty. DFO has low sensitivity to discrete rockfall impacts and struggles to distinguish a falling rock from passing trains, wind, or other vibration sources. Good coverage, poor confidence that a specific alarm corresponds to a specific rockfall. For continuous safety-critical alarming, that ambiguity is a problem.
LiDAR
LiDAR produces detailed 3D slope models and is a strong surveying and change-detection tool.
As a primary alarming system it has serious operational limits. Fog, rain, and snow scatter the laser; vegetation distorts readings; it needs frequent calibration; and it draws enough power that continuous operation realistically requires a mains supply, ruling out remote, off-grid, or alpine sites running on battery or solar. LiDAR belongs in survey and analysis, not in the real-time alarm loop.
Physical barrier-mounted detection
This approach mounts sensors directly on the protective barrier or fence — the exact point where a rockfall makes physical contact. Detection is mechanical and electronic rather than optical or line-of-sight, which removes most of the environmental ambiguity the technologies above struggle with.
Because the sensor sits where the impact happens, there is no question of whether a vibration “might” be a rockfall. The coupling between event and detection is direct. This is the technology class built specifically for real-time alarming rather than adapted from a surveying tool.
What “good” looks like in a railway rockfall alarming system
What a safety-critical system must deliver
- Speed: a few seconds, provenImpact-to-alarm measured in single-digit seconds, demonstrated in real deployments — not just data sheets.
- Redundancy with no single point of failureSensor coverage, network paths, and communication channels should all have backups.
- Independence from weather and visibilityPhysical detection works in rain, snow, fog, darkness, and vegetation. Optical systems do not.
- Operation in remote, off-grid terrainLow power draw with battery or solar support is non-negotiable for alpine and isolated sites.
- A real operational track recordYears of continuous service on a comparable mainline tells you more than any specification.
ImpactSentinel™: real-time alarming engineered for railway conditions
ImpactSentinel is a wireless rockfall and geohazard sensor built around exactly these requirements. It combines two complementary detection methods in a single unit: a MEMS module that continuously senses vibration, movement, and tilt, and a patented physical breakage mechanism that triggers an alarm once a defined force threshold is exceeded. Electronic sensing catches gradual movement; mechanical threshold detection catches sudden high-energy impact. Together they detect both slow slope failure and abrupt rockfall with high confidence, even in extreme conditions.
Mounted directly on the protective barrier, ImpactSentinel detects at the point of contact and delivers an alarm within 2–4 seconds — fast enough to support automated line closure and emergency braking. Its triple-redundancy architecture spans sensor coverage, network relays, and communication channels (RF, LAN, GSM, fibre, 3G/4G), so if any component fails, data reroutes automatically. Because detection is physical rather than optical, it is unaffected by rain, snow, fog, darkness, vegetation, or storms. Sealed to IP67, certified CE/UKCA, and rated from −20 °C to +60 °C, it is built for remote alpine deployment on battery or solar power.
When an event is detected, the system notifies designated personnel by SMS, updates a real-time monitoring portal, and drives on-site indicators — a three-tier warning system (green / orange / red) with local sirens for evacuation in serious incidents. Maintenance teams get an informed, immediate basis for action, even at sites with limited connectivity.
Proof: the Gotthard corridor, 10+ years in continuous service
ImpactSentinel is not a prototype. It has protected one of Europe’s most strategically important railway corridors for over a decade.
First installed at Gotthard tunnel portals in 2014, the deployment has grown into a network of more than 1,000 sensors across over 200 km of track through the Swiss Alps, in partnership with Swiss Federal Railways (SBB). The system was built to address a specific risk: high-speed trains exiting tunnels into rockfall-exposed alpine terrain with minimal driver visibility and reaction time. Across viaducts, deep valleys, and ridge crossings, it has proven low-maintenance and highly resilient, with a major sensor and firmware refresh in 2023 after nearly a decade of continuous operation.
The system’s resilience has been demonstrated under genuine extremes. During a severe storm and rockfall event along the Axenstrasse corridor in 2024, ImpactSentinel sensors mounted on the protective barrier functioned as the final decision layer in a multi-stage hazard system — recognising the event in real time and triggering the predefined safety protocol despite weather conditions that challenge optical and remote monitoring.
First deployed at Gotthard tunnel portals in 2014 — now protecting over 200 km of mainline track with more than 1,000 sensors.
— ImpactSentinel™ Gotthard Corridor Deployment
That is the difference between a system that records what happened and one that protects the train.
Specifying a rockfall detection system: where to start
If you are responsible for an exposed corridor, the practical path is straightforward:
Frequently asked questions
Monitoring collects data about slope movement over time for analysis and trend assessment. Alarming detects a specific event as it happens and triggers an immediate response — such as an SMS alert or automated line closure — within seconds. Monitoring tells you a slope is deteriorating; alarming tells you a rock is on the track right now. For safety-critical railway operations, real-time alarming is the essential capability.
For operational safety, impact-to-alarm time should be measured in single-digit seconds. A train travelling at 200 km/h covers more than 50 metres per second, so any delay of minutes renders an alarm useless for preventing an incident. ImpactSentinel delivers an alarm within 2–4 seconds of impact — fast enough to support automated line closure and emergency braking.
It depends on the technology. Optical systems (including LiDAR) and radar-based monitoring are significantly degraded by fog, heavy rain, snow, and vegetation. Physical barrier-mounted detection — where sensors are mounted directly on the protective fence — is unaffected by weather or visibility conditions, because detection is mechanical and electronic rather than line-of-sight.
ImpactSentinel has been in continuous operational service since 2014, first deployed at Gotthard tunnel portals in the Swiss Alps in partnership with Swiss Federal Railways (SBB). The deployment has grown to more than 1,000 sensors across over 200 km of mainline track, with a major refresh in 2023 after nearly a decade of uninterrupted operation.
Ready to protect your corridor?
INGLAS has protected mainline railway infrastructure in demanding alpine conditions for over a decade. Talk to our engineering team about the right approach for your highest-risk sites.
