PRESS RELEASE

An inconspicuous sensor on the rail, which detects any vibrations and excitements, sends a signal and artificial intelligence evaluates the potential risk. According to experts from the Faculty of Mechanical Engineering at BUT, this could be a recipe for increasing safety on the railway. They collaborate on research with domestic and Taiwanese partners. Following the recent rail accident in Taiwan, which killed more than 50 people and injured two hundred, safety has become a priority for the authorities. Therefore, the Taiwanese want to accelerate the test deployment of promising technology from the Czech Republic as much as possible.

On April 2, 350 passengers took the train from Taipei. An eight-car train derailed north of Hua-lien when it crashed into a truck crashing off the road at the mouth of a tunnel. The balance of the worst railway disaster in the history of the island was tragic. Similar tragedies could be helped by smart technologies, on which experts from the Faculty of Mechanical Engineering at BUT work.

"We will stick a sensor on the rail, which will become part of the rail, thanks to the electronics and the connection to the Internet of Things, it sends all captured signals about what is happening on the track to the cloud storage. We solve this part of the development in cooperation with the Czech companies ALIS Tech and Drážní revize, which manages the project on behalf of the Czech side. The Taiwanese side is in charge of signal processing using artificial intelligence methods. Should the outputs be connected to the rail systems, our system can serve as an early warning, for example, to stop a train. We could thus prevent similar accidents," Zdeněk Hadaš from the Department of Mechanics of Bodies, Mechatronics and Biomechanics, FME BUT describes.

Researchers originally developed smart technology as a complement to the current method of control, where the track must be visually monitored. "Originally, we intended the system for predictive maintenance applications, i.e. to make changes to the track before any damage is visible. But Taiwan also has other needs. It is a very mountainous island that is seismically active, there are often earthquakes or landslides. That is why it is interesting for them to use such smart measuring points for early warning, for example in combination with camera systems," Hadaš says, confirming that after the recent railway accident, Taiwanese partners are pushing for the system to be put into test operation as soon as possible. The development was financially supported by the Technology Agency of the Czech Republic and the Taiwanese Ministry of Science and Technology.

It catches every tap with a stone
Czech researchers have passed a series of first tests that show that the technology works. "We tried to detect any excitement from the sensor, just when we tapped a stone on the rail. If an unexpected situation occurs, for example, a landslide or rocks fall on the track, an earthquake breaks, the track breaks or a car hits it, we are able to capture this event and evaluate it on the server,” Hadaš explains. The range of the sensor is about 20 to 25 meters. "The technology is relatively affordable; we are able to make one scanning point for about 10,000 Czech crowns. The technology could be used immediately in critical places in the test mode," Hadaš assures.

The plans to install sensors in Taiwan are currently hampered by a coronavirus pandemic. Therefore, researchers are preparing further long-term tests in the Czech Republic, on the private tracks of two large industrial companies. "Ensuring long-term testing on a regular track with civilian traffic is difficult in the Czech Republic and in Europe from the point of view of legislation," Hadaš says. During the pandemic, they created a second-generation sensor with colleagues and also plan to develop their own components. "So far, we are buying commercially available piezoelectric composites, but if we develop our own, we can reduce the price of the system by an order of magnitude," Hadaš concludes.