Home » Indian Astronomers Discover Habitable Planets Using AI Tools

Indian Astronomers Discover Habitable Planets Using AI Tools

by admin
Indian Astronomers Discover Habitable Planets Using AI Tools

Artificial Intelligence (AI) has made significant progress in recent years, allowing us to advance in almost every field, including space exploration and technology. With the advancement of AI, scientists and researchers can solve complex problems such as clearing junk from Earth’s orbit and identifying life outside our solar system. Here’s proof from our very own Indian scientists and researchers. Indian astronomers have developed a new method for identifying potentially habitable planets based on artificial intelligence. The scientists were able to detect nearly 60 potentially habitable planets among 5000 other completely unknown planets because of this new algorithm.

 

What did the research reveal? 

 

Astronomers from the Indian Institute of Astrophysics, an autonomous institute of the Department of Science and Technology, Government of India, and astronomers from BITS Pilani, Goa campus, devised a brand new approach — an anomaly detection method — to successfully identify potentially habitable planets with a high probability of harbouring life in their recent study. “The method is based on the postulate that Earth is an anomaly, with the possibility of the existence of a few other anomalies among thousands of data points,” according to the study. The findings were reported in the journal Monthly Notices of the Royal Astronomical Society (MNRAS).

 

According to Indian astronomers, there are 60 potentially habitable planets among the approximately 5000 other planets confirmed to exist. The assessment was based on their close resemblance to Earth. “The fact that Earth is the only habitable planet among thousands of planets is referred to as an anomaly.” “We investigated whether similar anomaly candidates’ could be found using novel anomaly detection methods,” explained Dr Snehanshu Saha of BITS Pilani, K K Birla Goa Campus, and Dr Margarita Safonova of Indian Institute of Astrophysics. The habitability potential of a planet, according to researchers, necessitates knowledge of multiple planetary parameters derived from observations. This also necessitates astronomers using hours of expensive telescope time. As a result, manually scanning thousands of planets and then identifying planets that are potentially similar to Earth becomes a time-consuming task. This is where artificial intelligence can be put to good use in locating habitable planets.

 

How AI was used to discover the planets? 

 

Artificial Intelligence comes to the rescue.

 

Under the supervision of Dr Saha and Dr Safonova, Indian researchers created a novel AI-based algorithm to detect anomalies and extended it to an unsupervised clustering algorithm to identify likely habitable exoplanets (any planetary body that is outside the solar system and usually orbits a star other than the Sun) from existing exoplanet datasets.

 

The researchers’ AI-based method called Multi-Stage Memetic Binary Tree Anomaly Identifier (MSMBTAI) is based on a novel multi-stage memetic algorithm (MSMA). MSMA employs the broad concept of a meme, which refers to an idea or piece of knowledge that spreads from one person to another through imitation. 

 

The study has already identified a few planets that exhibit similar anomalous characteristics as our planet Earth using the proposed AI-based technique, and the scientists believe the results are promising. According to the astronomers, this will also aid in the future analysis of exoplanets and make the process much simpler.

 

Going into detail to learn more about this method

 

The name of this AI-based method is Multi-Stage Memetic Binary Tree Anomaly Identifier (MSMBTAI), which is also abbreviated as MSMA. This technique is based on the generic concept of a meme, which is an idea or knowledge that is passed down from one person to the next through imitation. Using the proposed technique, the researchers discovered a few planets that exhibited similar characteristics to Earth.

 

Detecting thousands of planets for life used to sound like a tedious task, but incorporating AI into the process has automated the difficult parts of the task.

 

Previous Research

 

In another study conducted in 2020, David Armstrong of the University of Warwick in the United Kingdom led a research team that was able to devise an algorithm that would efficiently determine which one is a planet, glitch, asteroid, and so on.

 

The data was then studied using NASA’s Transiting Exoplanet Survey Satellite. To put it simply, a drop in brightness indicated that something was passing by a star. It doesn’t always have to be a planet, but it did help narrow down the search. This prompted Armstrong’s research team to develop a machine-learning algorithm that was applied to NASA’s Kepler mission and was capable of distinguishing planets from other celestial objects. The algorithm was considered a massive breakthrough because it was capable of confirming up to 50 planets. “The algorithm we developed allows us to take 50 candidates across the threshold for planet validation, upgrading them to real planets,” Armstrong said in a Warwick release on August 25, 2020.


Also published on Medium.

You may also like