Introduction
Trains have relied on trackside signals for 150 years. GNSS is changing that, enabling positive train control, asset tracking, and more efficient operations across rail networks worldwide.
The Rail Environment
Challenges: deep cuts and tunnels (signal blocked), urban areas (multipath), remote areas (no cellular), and high speeds. Opportunities: predictable routes, known obstructions, and integration with precise track databases.
Positive Train Control (PTC)
Mandated in the US after fatal accidents, PTC prevents train-to-train collisions, enforces speed restrictions, and protects work zones. A GNSS receiver on the locomotive references a digital map of all track, speed limits, and signal states. The system intervenes if the engineer doesn't comply. Accuracy needed: <10 metres (sufficient for track identification).
Train Positioning Challenges
- Which track? Multiple parallel tracks require map matching and other sensors, GNSS alone is insufficient to determine the exact track
- Tunnels: Complete GNSS loss; odometry and inertial fill the gap; position resolved when exiting
- Urban canyons: Multipath from buildings; track database constraints help significantly
Multi-Sensor Fusion for Rail
A typical system combines a GNSS receiver, wheel odometers (tachometers), an IMU (inertial measurement unit), the track database, and balises (trackside beacons) that provide absolute position updates at known locations. A Kalman filter combines all these inputs for the optimal position estimate.
Asset Tracking
- Locomotives: Fleet management, utilisation tracking, maintenance scheduling
- Freight cars: Location, empty/loaded status, estimated arrival time
- Maintenance vehicles: Work crew locations, track inspection equipment, safety monitoring
GNSS combined with cellular is typical for asset tracking applications.
Track Inspection
Geometry cars measure track alignment with GNSS for location and inertial sensors for geometry, identifying defects precisely. Automated inspection by mounting sensors on revenue trains enables continuous monitoring and early problem detection. Drones inspect remote tracks with GNSS positioning for precise defect location.
Future: Autonomous Trains
Grades of automation range from GoA1 (driver with ATP) to GoA4 (fully autonomous). GNSS is the primary positioning source for most systems, combined with other sensors for integrity. Examples already operating: Rio Tinto autonomous trains in Australia, and numerous metro systems worldwide (Copenhagen, Paris).
Vital Points
- Positive Train Control uses GNSS for safety, mandated in the US
- Track identification requires map matching, GNSS alone isn't enough
- Tunnels and urban areas need sensor fusion
- Asset tracking improves efficiency across entire fleets
- Autonomous trains are a reality today on specific corridors
- Safety-critical integrity is essential throughout