RailEye® Front Collision Warning

Intelligent camera sensor for the detection of hazards

Warning for risk of collision

RailEye® Front Collision Warning

The RailEye® Front Collision Warning Assistant (FCWA) from EYYES recognises critical traffic situations with collision potential and actively warns the driver by means of visual and acoustic signals. The system is suitable for all rail vehicles and reliably warns of hazards. Thanks to artificial intelligence and the use of high-resolution camera sensors, the system only warns the driver if a dangerous situation actually exists. The system not only relies on detected objects in the front area of the vehicle, but also takes into account the current and pre-calculated direction of movement of the detected objects for a meaningful result.

Download Front Collision Assistant Folder

Technology

Functional description

If an object approaches the danger zone, this information is forwarded in real time via the data interface (Ethernet) to the central control unit (train computer) via an object list when a threshold value is reached that is derived from the calculated time until an imminent collision (Time to Collision).

Depending on the threshold values, this data can then be used to generate warning signals or, if the train driver is not paying attention, for automatic brake intervention.

Deep Learning at EYYES

Advanced object recognition

Visibility

With a sensor mounting height of 2.7 m and a mounting angle of 14°, the following detection characteristics result:

1.) Detected classes via DNN of the front collision warning system.

Passengers
Bicycles/Motorcycles
PASSENGER CARS
LORRY
Train

2) Detected objects via the track detection of the front collision warning system

Track closures
Buffer blocks

3) Detection range for straight track

50m* per class at ambient conditions according to VDV191
up to 80m* at favorable ambient conditions

* The actual minimum and maximum detection range depends on the possible positioning of the camera as well as the characteristics of the vehicle front and will be defined within the project.

Smart Components

Forward Facing Camera

Robust forward facing camera with excellent image quality
Advanced Driving Control Unit

Central control unit with integrated artificial intelligence for object recognition
Signaller

An optical and acoustic signal transmitter shows the status and reports actual dangers.

RailEye® Front Collision Assistant

Digital solution for the railway

With high-resolution camera sensors in combination with artificial intelligence
No need to use additional sensors such as radar or LiDAR
Actively informs the driver in critical situations by means of a 2-stage warning strategy
Calculation of the probable direction of movement of recognised objects
Only warns if there is an actual risk of collision by using several decision layers (desicion layers)
Detects people, cars, lorries, buses, trams, cyclists, obstacles on the track, e.g. buffer stops
Integrated self-diagnosis function and clear display of the system status

Increased safety and cost savings by avoiding accidents
Recognises the course of the track and thus the vehicle’s own direction of movement as the basis for defining the danger zone (area of interest)
Can be extended with Moving Off Assistant for checking the vehicle’s proximity when leaving the station area.
No need to break through the outer skin of the vehicle
Only one central computer unit, even for bidirectional vehicles!
Calibration, configuration and system updates, conveniently via web interface

Additional safety

RailEye® Moving-Off Assistant

The RailEye® Moving-Off Assistant monitors the near area of the front of the rail vehicle. The algorithm has been specially adapted to situations when entering and leaving stops and is a logical extension of the Front Collision Assistant for monitoring the area directly at the front of the vehicle.