Tunnels have unique challenges not found in other road conditions. These include low light, limited space to place cameras, distorted sound bouncing off the walls of the tunnel, constraints in servicing the equipment, equipment getting dirty from dust and other particles, etc. Manufacturers therefore need to come up with innovative solutions to ensure safety and security.
Tunnels have unique challenges not found in other road conditions. These include low light, limited space to place cameras, distorted sound bouncing off the walls of the tunnel, constraints in servicing the equipment, equipment getting dirty from dust and other particles, etc. Manufacturers therefore need to come up with innovative solutions to ensure safety and security.
The limited lighting options inside the tunnel poses as a challenge. To counter it, manufacturers use low light cameras and high dynamic ranges which can help counter the differences in lighting between the tunnel itself and its portals. Jenoptik, for example, deployed “Robot Black Flash” technology in two long Thuringian tunnels on German highways 4 and 71. Both tunnels require specific flash technology to ensure that drivers are not blinded by the flash. For that purpose, an infrared flash was developed which is almost invisible to the human eye and thus, cannot blind drivers, but nonetheless enables high-resolution image quality for the correct recognition of drivers and license plates.
Radar is used to detect vehicles regardless of weather or lighting conditions, and laser sensors are used inside the tunnel.
One of the challenges in tunnels is the large amount of environmental noise. This includes lighting changes, shadows, highlights and reflections. The video software must understand this type of noise and ignore it. “A second challenge with tunnels is low ceiling heights and constrained space, which result in camera mounting locations with very low look angles, but views which can reach to several hundred meters. At long distances and low look angles, cars appear to be merged and overlapped. The camera does not get to see the entire car until it is very close. Understanding the separation of vehicles in this situation is a challenging problem. At PureTech, we address this with the use of geo-referenced video. This provides us location, real size and speed information, which in turn, enables our software to recognize individual cars, versus something that looks more like a continuous train,” explained Larry Bowe, President and CEO of analytics provider PureTech Systems.
Other companies also employ the concept of pinpointing the exact location of vehicles inside the tunnel and its portals using other sensors, and not just video. Radar is used to detect vehicles regardless of weather or lighting conditions, and laser sensors are used inside the tunnel. The position coordinates of any found object are then usually transferred to a camera system to align the orientation of the camera towards the detected object.
The Main Risk Remains to be Fire
Fire is the main danger in tunnel security. Hence, it is essential to identify the cause of a problem in the tunnel or at the portals as early as possible. To enable this, modern tunnels combine several systems together for monitoring and alerting. Combining several systems gives better detection rates, and also redundancy and back up for different systems. “The combination of linear heat and fire detection systems and video-based automatic incident detection (Tunnel AID) ensures a very high incident detection rate with a very low number of false alarms in the tunnel,” said Urs Rufener, Security Applications Director at
Siemens Building Technologies. This system makes it possible to determine the precise location of a fire, and to activate automatic extinguishing even when smoke obscures visibility.
Combining several systems gives better detection rates, and also redundancy and backup for different systems.
“In the rare case of a fire inside a tunnel, it is also very important to know exactly the air (and so the smoke) velocity along the tunnel in order to control the ventilation system in an optimum way. For example, getting the smoke as a layer at the ceiling level and allowing people to run below the smoke to the next emergency exit. The knowledge about the air flow direction in tunnels is of course also necessary to plan the rescue procedures. Therefore, in many tunnels today, the air velocity and direction is monitored with three devices at the same place for getting full redundancy,” said Christoph Seewald, Head of Industry Management Transport Logistics for German sensor manufacturer SICK.
Fire detection also takes a proactive approach by trying to detect vehicles prone to catch fire. “In front of road tunnels with high safety requirements (e.g., very long ones with bi-directional traffic), in recent years there were some systems installed which automatically detect vehicles which might be about to catch fire. These systems do match a thermal image taken with high speed infrared cameras with 3D images – usually taken with laser scanners. The 3D images of the vehicles are analyzed in order to find a good segmentation (i.e., which are the axles, which are the exhaust pipe, where the engine is). Each segment has its own temperature limit. The hot spot detection system automatically finds then any vehicle with segments with too high temperatures. Based on this information, the dangerous vehicle can be guided to a nearby police station or something similar,” described Seewald.