IndigoVision IP technology improves passenger safety and speeds up operations for the Berlin S-Bahn

13 June 2007

IndigoVision is providing the technology behind a new train dispatch system that improves passenger safety and speeds up operations for the Berlin S-Bahn. The S-Bahn is a metro system that transports around 1.3 million passengers a day across a network consisting of 16 lines and 165 stations. This IP Video system allows the train driver to safely manage train dispatch via a CCTV display in the cab – the first installation of its type anywhere in the world for an urban rail system.

IndigoVision’s IP Video technology was chosen following a rigorous competitive analysis. This was followed by the installation of a pilot system using 24 cameras across 6 stations, which was successfully completed over a further 6 month period. The next stage will be to install a further 88 cameras on an additional 22 stations. In time the Berlin S-Bahn hopes to deploy the system across its entire network of 165 stations and stopping points. The system is part of an IP-communication and passenger information system installed by Alcatel-Lucent.

IndigoVision was chosen for this prestigious project because it could meet a stringent set of criteria. Even though the pilot was small the system had to be fully scalable to meet the future expansion plans, which could mean over 2,000 cameras monitoring 330 platforms across the entire rail network. Flexible video recording was also critical as was the ability to integrate alarm handling with video management, all of which is handled by IndigoVision’s Network Video Recorders and ‘Control Center’ software.

In a typical station, four fixed CCTV cameras are placed on each side of the track to monitor the platform and train doors. These four cameras are then connected to a quad-split unit to produce a single 4-way video image which is connected to an IndigoVision 8000 transmitter unit. The 8000 converts the analogue signal into a high-quality digital video stream, 4SIF at 25fps, for transmission over the IP network. The video is also transmitted using a wireless LAN to a display in the train cab where the driver can monitor passenger movements and ensure doors are closed before departure.

‘Control Center’, IndigoVision’s enterprise video and alarm management software is used by operators in control rooms, located in the larger stations, to monitor the platform images. The video streams are transmitted using multicast technology that allows any ‘Control Center’ workstation to view live and recorded video from any of the platform cameras at any station or stopping point on the rail network.

IndigoVision’s ‘Control Center’ software architecture allows the user to install as many operator workstations as required for no more than the cost of a PC. This will have a major impact on reducing costs as the system expands - another reason IndigoVision was chosen for the project.

The built-in motion analytics in the IndigoVision 8000 transmitters is used to trigger an alarm when a train arrives at the station by detecting motion in three quarters of the quad-split image. The alarm is used to update an interactive rail network status map displayed on ‘Control Center’ workstations and provide real time control of video recording.

Video recording is achieved using IndigoVision’s PC-based Window’s Network Video Recorders (WinNVRs). Three WinNVRs were installed with the original pilot and a further 4 are planned for the next phase of the project. The NVRs, which can be installed at any point on the IP network, continuously record video from all the platform cameras. Most recordings are deleted quickly but recordings of train movements are archived for later analysis.

This recording scheme is managed by ‘Control Center’ using the train arriving event triggered by the 8000’s motion analytics. The recording demands therefore change throughout the day and throughout the network. The distributed nature of WinNVRs operating on IP networks means that recorded video from both the larger central stations and the smaller stopping points can easily be accommodated and be fully scaleable as the system size increases in the future.