The Milky Way has 44 planets in other star systems that resemble Earth, according to a new machine learning model. The AI Earth-like planet predictor has been created by Researchers at the University of Bern and the National Centre of Competence in Research PlanetS (NCCR PlanetS), Switzerland.

Advertisment

They say that the algorithm employed by the model is 99 per cent accurate. Their Earth-like planet predictor had the ability to tell which planetary systems might have planets like Earth. The research was published in Astronomy & Astrophysics.

The scientists who were part of the study said that the AI model is extremely promising and can help astronomers focus their attention only on star systems that have a high chance of having exoplanets that might be able to host life.

Also Read: K2-18b: 'Revolutionary moment' as astronomers find 'strongest evidence' of life beyond our solar system

Advertisment

"The model identified 44 systems that are highly likely to harbour undetected Earth-like planets,” Dr Jeanne Davoult, lead author of the paper, said. 

The basis of discovering such exoplanets is the path of their orbit. A planet in the habitable zone is most likely to host life. This means it should be in a place where it is neither too hot nor too cold to be able to have liquid water. Exoplanets that fulfil this condition are the most promising candidates when looking for other worlds where life can exist.

The AI model was tested using known data about other planetary systems. The authors said that when information about systems whose properties the scientists are aware of was fed to the model, it did amazingly well. 

Advertisment

According to the AI model, there are 44 such planets in the Milky Way. The researchers tested the results further, and a second study proved that there is a "theoretical possibility for these systems to host an Earth-like planet."

AI model is 99% accurate

Davoult says the model achieves precision values of up to 0.99. This means that of all the results, at least 99 per cent of systems have at least one Earth-like planet.

"This is a significant step in the search for planets with conditions favourable to life and, ultimately, for the search for life in the universe," co-author Dr. Yann Alibert said.

NASA, ESA and other space agencies are regularly looking to discover exoplanets that might be similar to Earth. Testing every planetary system is not possible. An AI model like this one can prove to be of great help in the endeavour by telling astronomers where exactly to look.