A PLAN has been unveiled to save Earth from an apocalyptic asteroid strike using artificial intelligence (AI).
Planetary defence experts said they have a number of options to protect mankind from being wiped out by a rogue space rock.
The three proposals include firing a nuke at it, smashing another object into it or changing its course using the gravitational pull of a spacecraft.
But researchers are not able to actually test each method – which is where AI comes in.
Speaking this week at a meeting of the American Astronomical Society, Erika Nesvold, an astrophysicist at the Carnegie Institution for Science in Washington DC, said: “We obviously don’t have the money or time to develop and test all of [the methods].
“So we want to know how can we decide which of these technologies is most likely to work.”
As part of their project – dubbed Deflector Selector – her team simulated more than 18 million different attempts to save Earth from a killer asteroid.
These results were then used to teach a computer to figure out the best solution to stop the asteroid.
After completely inputting all the data into the computer, it was able to decide how to stop the potentially apocalypse-causing asteroid in seconds.
Nesvold’s team is hoping that by further improving the AI system, they can fully protect the planet.
NUKED: One of the proposals to stop an asteroid is to nuke it
Just weeks ago, NASA revealed Asteroid 99942 Apophis– a rock as big as a mountain – could hit Earth in 2036.
When first discovered in 2014, Apophis had a record-breaking collision risk of four on the Torino scale.
Steve Chesley, a NASA scientist, along with Paul Khodas from the space agency’s jet propulsion lab, have now predicted impact will occur on April 13, 2036.
“Apophis has been one of those celestial bodies that has captured the public’s interest since it was discovered in 2004,” said Chesley.
“Updated computational techniques and newly available data indicate the probability of an Earth encounter on April 13, 2036, for Apophis has dropped from one-in-45,000 to about four-in-a million,” he added.