How AI is supposed to increase safety in trams - Bremen Deploys AI to Tackle Tram Conflicts in Groundbreaking Pilot

Bremen's public transport network has introduced an AI-powered safety system to reduce conflicts on its trams. The technology, rolled out as a nationwide pilot in April 2023, detects aggressive behaviour in real time. No other German city currently uses a similar AI-driven approach for monitoring public transport.

The system works by analysing live footage from tram cameras. It instantly flags rowdy conduct, verbal clashes, or physical threats. Before processing, the AI automatically pixelates passengers' faces to ensure anonymity and comply with data protection laws. State regulators have already approved the measures.

When the AI identifies a potential incident, alerts are sent to drivers and control centres. Staff can then assess the situation via video feeds and, if needed, summon emergency services. The system was first installed on 15 trams in April 2023, with plans to expand to a third of Bremen's fleet by the end of the year. Eventually, every tram and bus in the city will feature the technology.

The move follows persistent safety concerns on Bremen's network. Officials report around 150 disputes annually, including 90 verbal attacks and 30 to 40 physical assaults. The AI aims to curb these incidents by enabling faster responses to emerging threats.

The AI safety system marks a first for German public transport. By the end of 2023, it will cover a significant portion of Bremen's trams, with full implementation planned across all vehicles. The technology is designed to reduce conflicts while adhering to strict privacy and data protection standards.