Deutsche Bahn is Germany’s national railway operator and one of Europe’s largest mobility and logistics companies, serving millions of passengers every day.

DB Fernverkehr, its long-distance passenger division, runs high-speed ICE trains across an extensive network where safety, reliability and on-time performance are paramount.

Visit Deutsche Bahn

Every year, Germany’s Federal Railway Accident Investigation Board records hundreds of SPAD incidents (Signal Passed at Danger events) where a train crosses a stop signal without authorisation. These incidents are growing at roughly +3% per year, disrupting operations, generating significant follow-up costs and, most importantly, creating real safety risks for passengers and crew.

Today’s primary safeguard, the PZB system, only intervenes after a driver has already passed a signal, by which point precious seconds and metres have been lost. The gap in protection lies upstream: in the critical seconds before the signal, where human perception alone stands between a routine journey and a dangerous incident.

DB Fernverkehr asked Sentigrate to investigate the feasibility of a complementary, software-driven early warning system — one that supports drivers with real-time, continuous feedback as they approach stop signals, without requiring any new hardware or trackside infrastructure.

The approach Sentigrate worked out consists of three steps:

A fixed GPS sample every ~10 seconds is far too coarse to warn a driver in time. By adapting the sampling rate as a train approaches a stop signal down to 2.8 seconds, the system keeps its proximity estimate accurate exactly when it counts.

The feasibility study confirmed the concept works with data already available on board, clearing the path from study to deployment.