AI Tool Validates Over 100 New Exoplanets in NASA TESS Data

Researchers have used artificial intelligence to analyze NASA’s TESS data, resulting in the validation of over 100 new exoplanets. This development highlights the growing role of machine learning in the identification and confirmation of planets beyond the solar system.

The RAVEN algorithm, created by a team at the University of Warwick, is designed to process and validate transiting exoplanet candidates found in TESS observations. The system applies machine learning models trained on simulated data to classify potential planets and uses statistical methods to confirm their status.

RAVEN’s process includes detecting transit signals with a box least squares algorithm, followed by classification and statistical validation. The analysis utilized TESS full-frame images from sectors 1 to 55, which cover the satellite’s first four years of operation.

What the numbers show

• 118 exoplanets were validated using RAVEN, with 31 newly detected in this analysis
• Over 2,000 vetted candidates were identified, including about 1,000 high-probability planets
• TESS sectors 1–55, representing four years of data, were used in the study

In addition to the validated exoplanets, the RAVEN pipeline produced a list of more than 2,000 candidates that have not yet been confirmed. Among these, approximately 1,000 are considered high-probability planet candidates based on the algorithm’s assessment.

Prior to the development of RAVEN, NASA’s Ames Research Center had introduced the ExoMiner AI model, which identified thousands of exoplanet candidates from TESS and Kepler data. ExoMiner was responsible for validating 370 new exoplanets from the Kepler mission, demonstrating the effectiveness of artificial intelligence in this field.

Media coverage has reported that the RAVEN algorithm uncovered over 100 new exoplanets previously hidden within NASA’s TESS data. These findings were attributed to the application of AI developed at the University of Warwick.

The use of machine learning tools like RAVEN and ExoMiner is expanding the catalog of known exoplanets by efficiently processing large volumes of observational data. These methods allow researchers to identify and validate planets that might otherwise remain undetected using traditional techniques.

As more data from TESS and other missions become available, the continued application of AI-driven pipelines is expected to support ongoing efforts to discover and confirm new exoplanets. The results from RAVEN and similar tools illustrate the potential for artificial intelligence to enhance astronomical research.

* This article is based on publicly available information at the time of writing.