A Transport for London AI-megfigyeléssel teszteli az utasok viselkedését
Transport for London (TfL) recently trialed the use of an AI surveillance system within the Willesden Green station. The system was applied to every camera in the building to spot dozens of different incidents. The trial results show the promise of AI-powered public services for safety and efficiency, while also raising concerns about the potential for constant, unblinking surveillance.
- The system could identify up to 77 different 'use cases,' though only 11 were used during the trial.
- Capabilities ranged from detecting fare evasion and crime to identifying spilled drinks or discarded newspapers.
- Safeguarding features alerted staff if a person sat on a bench for longer than 10 minutes.
- Alerts were also triggered if a person remained in the ticket hall for more than 15 minutes.
- The trial used existing camera infrastructure to provide always-on monitoring.
Miért fontos?
AI lets you have an unblinking, always-on surveillance capability for anything you can specify. This is largely beneficial for public safety, but it also provides a blueprint for Orwellian surveillance. The trial demonstrates that mundane computer vision applications have made the leap from research papers into production systems.