Artificial Intelligence Detects 72 Mysterious Radio Bursts from Space

Artificial Intelligence Detects 72 Mysterious Radio Bursts from Space

FRB 121102 emitted 21 bursts previously detected during Breakthrough Listen observations made in 2017 with the Green Bank Telescope in West Virginia.

Using the latest machine learning techniques, the programme found the FRBs emanating from a "repeater" called FRB 121102.

But it is still unclear what is actually causing them: suggestions have included everything from highly magnetised neutron stars to messages being created by alien technology.

FRBs are typically one-time events - which makes FRB 121102 particularly interesting as it's given off hundreds of bursts.

FRBs are known to be millisecond pulses of radio emission from galaxies far away. The blasts, which last just a fraction of a second each, are impossible to forecast, but because this particular spot in the sky has produced so many in the past it's easy for astronomers to assume that more might be coming. The AI pointed to 72 additional flashes, bringing the total to 93 fast radio bursts from FRB 121102 on that single day.

Next, UC Berkeley Ph.D. student Gerry Zhang and a few collaborators developed a new, powerful machine-learning algorithm - using similar techniques implemented to optimize search engine results. The 21 fast radio bursts were all seen within one hour, which suggests that whatever the source of FRB 121102 is, it demonstrated a period of excessive activity.

"Gerry's work is exciting not just because it helps us understand the dynamic behavior of FRBs in more detail", said SETI's Dr Andrew Siemion, "but also because of the promise it shows for using machine learning to detect signals missed by classical algorithms".

More news: Sweden faces weeks of uncertainty after close election

The initiative may advance the search to find signs of intelligent life in the universe, said researchers from the University of California, Berkeley in the US.

A group of scientists in the framework of the project a Breakthrough Listen analyzed the signals received from an unknown distant source.

As a January report in The New York Times noted, "Among the more out-there explanations proffered was that they are lasers propelling alien interstellar spacecraft". They trained an algorithm known as a convolutional neural network to recognize bursts found by the classical search method used by Gajjar and collaborators, and then set it loose on the 400 TB dataset to find bursts that the classical approach missed.

Breakthrough Listen is also applying the successful machine learning algorithm to find new kinds of signals that could be emerging from an extraterrestrial civilization.

The Listen science team, which, after a five-hour long observation led by Gajjar a year ago had detected the FRB-121102 in 2017, has now developed a new, powerful machine learning algorithm which has reanalysed the 2017 dataset which has led to the 72 new FRBs.

The new algorithm was very helpful in determining that source FRB121102 does not send out bursts at regular intervals (or at least not intervals longer than about 10 ms).