Mystery Signals Coming From Space, Scientists Discover

( Late last month, a group of astronomers reported detecting eight “suspicious” radio signals that could be evidence of “technological life beyond earth.”

Led by University of Toronto student Peter Ma, the team of astronomers created an artificial intelligence algorithm that enabled them to detect the signals while examining 820 stars using the Green Bank Telescope located in West Virginia.

Using machine learning, the algorithm was able to differentiate between man-made signals like those from cell phones and GPS satellites and those that were potentially extraterrestrial in origin, allowing them to detect eight suspicious signals that could not be detected during previous observations at the Green Bank Telescope due to interference.

In a paper published in the journal Nature Astronomy in late January, Ma explained that while the eight radio signals were not definitive proof of extraterrestrial life, the unexplainable nature of the signals will likely fuel speculation that there is life beyond Earth.

Scientist Steve Croft, who also worked on the project, said radio techno signature searches are much like looking for a needle in a haystack with the “vast majority” of signals detected by telescopes originating from “our own technology.”

Given the “narrow band” of the eight radio signals, Croft believes they could come from an extraterrestrial source since man-made signals are generally broadband. What’s more, the eight signals contained a “slope,” which indicates that their origin was accelerated with antennas that are unlikely to have come from Earth.

According to Peter Ma’s research advisor Cherry Ng, the results of the project provide a dramatic illustration of how machine learning can be applied to “data challenges in astronomy.” Ng added that applying the techniques on a larger scale would be “transformational” for the radio techno signature science.

Peter Ma hopes to expand the use of the AI algorithm to examine a million stars using the MeerKAT telescope in South Africa. He said the machine-learning technique could help “accelerate the rate” astronomers can “make discoveries” in the effort to determine whether or not humans are “alone in the universe.”