At a Glance
- A high school student from California, Matteo Paz, developed an AI model that uncovered 1.5 million previously unknown objects in space, improving NASA’s NEOWISE telescope data analysis.
- Paz’s AI model processed data from the NEOWISE telescope, which has tracked asteroids and variable cosmic objects like quasars and exploding stars, identifying them with higher accuracy than traditional methods.
- Using machine learning, Paz trained the AI model on large datasets, allowing it to detect small changes in infrared measurements and successfully catalog millions of previously unnoticed space objects.
- This breakthrough in space data analysis opens the door to discoveries and demonstrates how AI can be applied to other time-based studies, such as financial charts and environmental monitoring.
- Paz’s work showcases the potential of young researchers to revolutionize data analysis across various fields, pushing the boundaries of science and technology through innovative applications of artificial intelligence.
A high school student from California has made a groundbreaking contribution to astronomy by developing an artificial intelligence (AI) model that uncovered 1.5 million previously unknown objects in space. Matteo (Matthew) Paz, a Pasadena Unified School District student, worked with astronomers at Caltech to improve how NASA‘s NEOWISE telescope data is analyzed. His discovery, whose results were published in The Astronomical Journal, could help astronomers gain deeper insights into the dynamic objects of our universe.
The AI model Paz developed was designed to process massive amounts of data gathered by the NEOWISE telescope, which has been observing space for over 10 years. The telescope focused initially on tracking asteroids and near-Earth objects and collected information on cosmic objects that show variable behaviors, such as fluctuating brightness. These variable objects, including quasars and exploding stars, are challenging to identify with traditional methods, but Paz’s AI was able to spot them with high accuracy.
Paz’s approach involved using machine learning, a type of AI that learns patterns in data. With guidance from his mentor, Dr. Davy Kirkpatrick, at Caltech, Paz trained the model on large datasets, allowing it to detect small changes in the infrared measurements from the NEOWISE telescope. After processing the data, the model successfully identified and cataloged millions of objects that previously went unnoticed, offering a wealth of new targets for future research in astronomy.
This research marks a significant leap in analyzing space data and could open doors to discoveries in various fields. Paz’s work on this AI-driven analysis has contributed to the study of space and demonstrated how AI can be applied to other time-based studies, like financial charts or environmental monitoring. His work is a powerful example of how young researchers can push the boundaries of science and technology, potentially revolutionizing how data is used across multiple disciplines.
References
- Paz, M. (2024). A submillisecond fourier and wavelet-based model to extract variable candidates from the neowise single-exposure database. The Astronomical Journal, 168(6), 241. https://doi.org/10.3847/1538-3881/ad7fe6
- Motrunich, A. & California Institute of Technology. (2025, April 11). High school student uses AI to reveal 1.5 million previously unknown objects in space. Phys.Org; California Institute of Technology. https://phys.org/news/2025-04-high-school-student-ai-reveal.html
