Stretching out thousands of miles beneath our feet, our fibrous ears listen. When you walk or drive over buried optical fiber, ground activity creates unique vibrations that slightly disrupt the path that light travels down the cable. With the right equipment, scientists can analyze the disturbance and determine what the source was and when exactly it was wandering there.
This rapidly growing technology, known as distributed acoustic sensing (DAS), is so sensitive that researchers recently used it to monitor the cacophony of cicada mass emergence. Did. Some companies are using the cables as ultra-sensitive equipment to detect volcanic eruptions and earthquakes. Unlike traditional seismometers that are fixed in one location, the network of fiber optic cables covers the entire terrain, providing unprecedented detail of ground rumbling in different locations.
Scientists are currently experimenting with bringing DAS to a train near you. When a train travels over a section of track, it creates vibrations that analysts can monitor over time. If that signal suddenly changes, it could indicate a problem with the rail, such as a crack or a broken sleeper. Or if a landslide explodes across the tracks on a mountain pass, DAS might also “hear” it and alert the railway operator to a problem that was not yet noticed by the human eye. If the signal changes more slowly, defects in track alignment can occur.
Coincidentally, fiber optic cables are already being laid along many railways to connect all signaling equipment and for communications. “We can reduce costs because we use equipment and infrastructure that is already available for this,” says engineer Hossein Taheri. study Georgia Southern University’s DAS for Railroads. “Some railroads may not have fiber, so you have to lie down. But yes, most of them usually already have it.”
To use that fiber, you need a device called an interrogator, which fires laser pulses into the cable and analyzes the tiny bits of light that bounce back. So suppose a stone hits a railroad track 20 miles from the interrogator. This creates characteristic ground vibrations that disturb the optical fibers near the tracks and appear in the optical signal. Because the scientist knows the speed of light, he can accurately measure the time it takes for the signal to return to the interrogator and pinpoint the distance to the obstacle to within 10 meters, or about 30 feet. Masu.
For a given track section, the DAS signals should have been analyzed over time to build a normal and healthy railway vibration profile. If your DAS data suddenly starts showing something different, there may be a problem, just like an EKG detects a problem with a human heartbeat. “What we do is profile the track and look for changes in its acoustic signature,” says Daniel, a railway expert and spokesperson for Sensonic, which develops his DAS technology for railways. Pyke says. “I know which truck it is.” should Like I know what a train is should Sounds like. And if that’s changing, for example if this joint is loose, we know that someone needs to go and fix it before it becomes a problem. ”
Pike said Senseonic’s system can monitor a trajectory of 40 kilometers (25 miles) in both directions from the interrogator. He added that having this type of system in continuous operation could reduce the human labor needed to inspect railway tracks around the world. This inspection is a dangerous job considering that there are huge machines flying around. If someone is digging cables to sell copper, Sensonic can detect that too. Sensonic can also detect when people are just walking and trespassing along the railroad tracks.
Even stranger, in India, SenSonic detects elephant footsteps near railway tracks to protect elephants and train passengers. An alarm would then sound to alert staff to a potential collision. “We actually had to hire elephants to roam the tracks,” Pike says. “This was one of the most interesting expenses he has ever submitted.”
The challenge is that DAS generates most of the data. too much lots of data. A single sensor is not installed at one point along the track, but extends over a vast distance above and below the rail. That means data comes in day and night over fiber optic cables from 40 meters away to 40 kilometers away, and everywhere in between. “The file you generated is HugeSo we need to use machine learning to automate that,” says David Milne, a research engineer at the University of Southampton. study DAS and railways. “There’s going to be an enormous amount of data. I don’t think it would be manageable or economical unless you had a computer to help you with it.”
Sensonic says it trained its AI on real railway data to recognize events such as falling rocks amidst all the noise. And the alerts sent to rail operators are only a kilobyte in size. “The machine learning and AI models used to identify these events are continually being improved to improve sensitivity and reduce false alarms,” Pike said.
The use of DAS for various applications, including railways, is still in its infancy, so researchers are still refining these systems. “Distributed acoustic sensing is one of the areas suppliers and carriers are considering to see if it can meaningfully advance their safety goals,” said Jessica Kahanek, a spokeswoman for the Association of American Railroads. states. “When railroads test new technology, they are not only looking to see if it works in the lab, but also to see if it works when exposed to the harsh operational realities of an outdoor network spanning continents. ”
No matter the use case, we’ll likely hear a lot more about DAS in the coming years, as DAS technology “listens” to the ever-increasing amount of disturbance on the ground.
