To develop the new technology, scientists from University of California – Berkeley recorded radio data from a source of cosmic radio energy called FRB, which is a collapsed and highly magnetized neutron star. The radio signals from data were recorded over a five-hour period on Aug. 26, 2017, by the Green Bank Telescope in West Virginia.
Using data from this source, the machine learning algorithm learned new patterns and subsequently uncovered 72 new bursts. These newly detected bursts would have gone undetected using conventional technology. The artificial intelligence is now equipped to probe deeper into the cosmos to look for new signals.
Fast radio bursts are bright pulses of radio emission. They have been hard to detect, lasting for just a few milliseconds. The source of such radio emissions is uncertain. While some are traceable to gas streams coming from neutron stars, some could be indicative of technology signatures relating to advanced civilizations.
The new initiative is called Breakthrough Listen and it represents the biggest ever astrophysics research program. One of the aims is to seek evidence of sophisticated alien life.
The program includes a survey of the 1,000,000 closest stars to Earth. This is undertaken by technology scanning the center of the Milky Way and then across the entire galactic plane. Beyond our galaxy, the program can listen for radio messages that might emanate from the 100 closest galaxies to ours.
The monitoring instruments used for this purpose are 50 times more sensitive than existing telescopes and the radio sensors can cover ten times more of space than any previous programs, and assess five times more of the radio spectrum. In terms of sensitivity, the sensors are said to be 1000 times more effective at finding laser signals than any other visible light surveys.
The new initiative will last for ten years and the budget for this period has been set at an astronomical $100,000,000.
According to one of the researchers, Dr. Andrew Siemion: “This work is exciting not just because it helps us understand the dynamic behavior of fast radio bursts in more detail, but also because of the promise it shows for using machine learning to detect signals missed by classical algorithms.”
The new research has been published in The Astrophysical Journal. The paper is titled “Fast Radio Burst 121102 Pulse Detection and Periodicity: A Machine Learning Approach.”