Sensors & Measurement — What the railway knows about itself
Sensors and measurement systems form the observational layer of the modern railway, enabling operators, infrastructure managers, and maintainers to monitor the physical state of rolling stock and fixed infrastructure continuously — replacing periodic manual inspection with automated condition assessment at a scale and frequency not achievable by human observation alone.
A railway network is a geographically distributed system with thousands of assets — tracks, switches, bridges, vehicles, overhead lines, substations — that degrade continuously and in ways that are not always visible between scheduled inspections. Sensor systems provide near-continuous observation, closing the gap between inspection intervals and enabling intervention before failure causes service disruption.
Rail sensing divides into two distinct domains: the vehicle and the infrastructure. Each requires different hardware, installation environments, and data architectures.
Onboard sensing
Sensors mounted on or within rolling stock monitor the mechanical and electrical state of the vehicle during operation. Accelerometers measure vibration in bogies, axle boxes, and carbodies; temperature sensors track bearing conditions, motor windings, and brake disc surfaces; pressure sensors monitor pneumatic braking circuits, suspension systems, and door mechanisms.
The onboard environment is demanding: sensors must tolerate continuous mechanical vibration, wide temperature ranges, electromagnetic interference from traction equipment, and exposure to moisture, dust, and debris. Hardware is specified to EN 50155 for electronic equipment on railway vehicles, along with applicable electromagnetic compatibility (EMC) requirements.
Data from onboard sensors is transmitted via the train’s internal network — governed by IEC 61375 (Train Communication Network) — to an onboard data recorder, a central train management system, or a remote condition monitoring platform.
Wayside sensing
Wayside systems monitor infrastructure and passing trains from fixed positions along the route. Wheel impact load detectors (WILDs) identify flat spots and out-of-round wheel conditions as trains pass at line speed; acoustic emission sensors at bearing monitoring stations detect incipient bearing defects without requiring the train to stop.
Wayside monitoring enables detection of rolling stock defects without removing the vehicle from service, making it a cost-effective complement to onboard sensing. Track geometry measurement captures rail profile, alignment, cant, and gauge across the network — from dedicated geometry measurement vehicles or from sensor arrays integrated into in-service trains.
Catenary and pantograph monitoring systems assess current collection quality at speed using cameras, contact force sensors, and wire position sensors. Structural health monitoring systems instrument bridges, tunnels, and embankments with strain gauges, inclinometers, and displacement sensors.
The data chain
A sensor alone produces no operational value. Value emerges from the full chain: sensor — signal conditioning — transmission — storage — analysis — action.
Signal conditioning converts raw sensor output into a usable digital signal. Transmission routes data to edge processing onboard or offsite via GSM-R or Future Railway Mobile Communication System (FRMCS) links, where rule-based alarming or statistical modelling identifies conditions requiring intervention.
The balance between edge processing — where algorithms run onboard or wayside in real time — and cloud-based batch analysis is determined by latency requirements, data volume, and connectivity constraints.
Condition-based maintenance
The primary operational application of rail sensing is condition-based maintenance (CBM), which replaces fixed-interval maintenance schedules with maintenance triggered by measured asset condition. CBM reduces unnecessary interventions, extends asset service life, and lowers the risk of in-service failures.
The shift from time-based to condition-based maintenance is a central objective of the EU’s Shift2Rail and Europe’s Rail Joint Undertaking (EU-Rail) research programmes, both of which have funded sensor integration work across European network operators and rolling stock fleets. Asset management platforms — covered in the Digital Systems & Software section — connect sensor data to maintenance planning systems.
Standards and certification
Safety-critical sensing applications require certification under EN 50129, which governs safety-related electronic systems for railways and defines Safety Integrity Levels (SIL 1–4). Condition monitoring sensors for maintenance applications typically operate under lower regulatory requirements.
Track geometry measurement is governed by EN 13848, which defines quality index parameters and intervention thresholds across line categories and speed ranges. Compliance with EN 13848 thresholds is required for infrastructure managers under ERA oversight within the European Union.
Six articles, one system
Article 2.1 covers accelerometers and inertial sensors — the most widely deployed vibration measurement technology across rolling stock and infrastructure applications. Article 2.2 addresses track geometry measurement systems and the standards that govern them.
Article 2.3 examines wheel and axle monitoring, covering both wayside detection systems and onboard approaches. Article 2.4 covers pantograph and catenary monitoring systems.
Article 2.5 addresses structural health monitoring of fixed infrastructure. Article 2.6 covers pressure and temperature sensors and their principal applications in rolling stock systems.