Railways: AI maintenance of railway crossing | AeyeQ
Case study

The safety of railway crossings is a challenge. A place where multiple traffic agents (cars, trains, and pedestrians) meet should have stricter control over alarm signals and equipment that regulate vehicle movements.

Even with the alarming and safety equipment, 5000 accidents happen in railway crossing in a year. The most frequent reasons are traffic violations and malfunctioning safety equipment. This part of the road is monitored by CCTV cameras and an employee who is responsible for many crossings at once. 

AI maintenance of railway with AeyeQ

AeyeQ has developed a smart system to AI automate preventive maintenance of safety equipment and discourage hazardous behavior of transport to solve this problem and help railway companies to increase the productivity of their CCTV camera. This is an ongoing project that was successfully implemented in 2021 with a smart vision system installation to perform the following functions:   

Gather anonymous statistics about the traffic

An AI-powered vision system produces statistical data about roads capacity to investigate the traffic situation and improve the movement of vehicles. The cameras not only count the participants but distinguishes between types of vehicles and pedestrians.   

Perform preventive maintenance of railway equipment

AeyeQ cameras ensure all preventive measures are taken to safeguard the crossing’s traffic. The smart system timely identifies broken barriers, malfunctioning alert lighting, or/and bells sounds and gives an immediate reaction to the control room. It distinguishes between the different barriers' positions, correct sound, and light alarm. In addition, our smart vision system detects damaged road parts and accidents involving railway barriers. 

Monitors hazardous behavior

Not all crossing participants obey traffic rules. AI cameras are trained to recognize risky behavior such as passing on a red light or in the chicane, catch movement exceeding 30km/h speed, and more to mitigate behavioral and traffic unpredictability. Timely detection of those problems positively contributes to the general safety of railway crossing and reduces the risk of further incidents.

All detections operate on the edge. It means that all calculations occur on the camera device. AI devices detect objects in real-time and send an instant response to the control room.  

Our smart vision system gives advantages: 

- Decrease numbers of accidents  

- Discourage malicious and risky behaviors  

- Perform preventive maintenance of safety and alert equipment quicker   

- Create statistics about traffic and road capacity 

How to choose the right camera for deep learning software?
Figure out how to find the perfect vision hardware for deep learning solution