Detection of
Violence
The AI detects clear physical altercations in real time and enables intervention before they escalate.
Sector
AI Watch on trains and buses
Vehicle interiors are enclosed, confined spaces. That is exactly where conflicts escalate quickly and where help often takes longest to arrive. Galvorn makes dangerous situations visible in real time and alerts drivers and the control room when fast help counts.
What Galvorn can do
Detection of
Emergencies
If someone falls badly and stays down, a report goes to the control room.
Detection of
Vandalism
Serious criminal damage is reported in real time, not at the next inspection round.
Automatic
Passenger counting
Occupancy is recorded continuously as a basis for billing and planning.
Real-time hazard alerts
All scenes are staged test recordings, not real incidents.
Easy to retrofit
We work with the cameras already installed in your vehicles. Galvorn turns them into hazard detectors without costly camera replacement. Easy to maintain, future-proof and powerful.
Existing cameras
Compatible with all common IP camera systems. Galvorn is connected over the network and is ready to use.
Control room connection
Alerts are transmitted directly to the control room. Optionally: notification of the driver and/or the police.
Scalable
Galvorn scales from a single train to the entire network. Central management via the control room is possible.
Intelligence for every train
Our AI runs directly on the train or bus, offline and secure. The hardware required is designed specifically for rail use and can be integrated into the vehicle’s systems within minutes.
In the Vehicle
No internet connection required. Compatible with 24 V on-board power. The system runs fully self-sufficiently inside the vehicle.
Certified
EN 50155:2021 for electronics in rail vehicles. EN 45545-2 for fire protection in rail vehicles.
One system per train
Powerful AI models thanks to 85 TOPS and 64 GB of graphics memory for up to 12 cameras with real-time analysis.
The same hardware can also be used for our driver assistance and predictive maintenance solutions. More information on the website of our sister company Bremen Robotics.
Implement with funding
There are suitable options for your project, too. Funding programmes for security technology and passenger data in public transport are available from transport associations, the federal states, the federal government and the EU. Some examples:
Transport associations
In North Rhine-Westphalia, associations such as go.Rheinland fund retrofitting buses and trains with passenger counting systems.
Federal states
Baden-Württemberg covers up to 75 % of the investment in automatic passenger counting systems and also subsidises ongoing operation. Saxony funds security technology in vehicles and at stops.
Federal government
BSAG introduced AI Watch with 80 % funding from the federal transport ministry programme “Digitalisation of municipal transport systems”.
EU
EU funds are an option, too: in southern Lower Saxony, 128 buses were fitted with counting systems through the “Zukunftsregionen” programme.
We work out with you which programme fits your project and provide the technical documents for the application.
Request a funding checkReferences & press
AI Watch has been running on the trams of Bremer Straßenbahn AG (BSAG) since 2025. The system was developed together with BSAG and the Bremen police, tested in live operation and is now being rolled out across the fleet step by step.
AI surveillance on the tram: more safety in the passenger area?
In-depth video report by heise online on AI Watch in Bremen’s trams, including interviews with Galvorn / JUST ADD AI and BSAG.
Watch the video
Security in public transport: how AI is meant to make Bremen’s trams safer
TV report by buten un binnen on threats and violence on Bremen's trams and the AI the BSAG is using to make tram travel safer.
Watch the video
Pilot project in Bremen: AI to make trams safer
After a year of trial operation, BSAG is now rolling out AI Watch: 17 trams are equipped, 43 are planned by the end of the year, with 80 per cent funded by the federal government.
Watch the video