Our Solution:
We installed our Situate technology on Platform 11 at BTM station. The system was equipped with two newly developed algorithms designed to detect unsafe behaviours:
- Yellow Line Crossing Detection: Alerts when passengers step beyond the yellow safety line for more than five seconds.
- Running Detection: Identifies individuals running on the platform and issues an audio alert advising them to slow down.
In addition, we built a new speaker system installed with our sensors, allowing us to channel alerts directly to passengers. For instance, asking them to ‘step back’ if they are over the yellow line. To make this effective, we built in live train detection so that alerts could be triggered only if the platform was empty.
We further enhanced the system with a cloud-based database that processes data in “super real-time”, ensuring minimal latency and delivering alerts when passengers and station staff need them most.
The Results:
Our trial demonstrated the effectiveness of Situate in improving platform safety:
- Passengers received immediate feedback on unsafe behaviour, reducing risks.
- The yellow line detection algorithm achieved 79% accuracy, while the running detection algorithm reached 65% accuracy.
- System successfully alerted when the platform was empty, but not when a train was in the platform (when passengers should be crossing the yellow line).
- Real-time data allowed station staff to monitor crowding trends and assess the total numbers of unsafe instances.
- Encouraged by the success of the technology, Network Rail has extended the trial to a year-long deployment in Kings Cross station to address crowd management challenges.
