Imagine turning a simple summer project into a discovery that stuns professional scientists. That\u2019s exactly what happened when a California high schooler uncovered over 1.5 million unidentified space objects hidden in NASA\u2019s abandoned data. What began as curiosity and a love for astronomy has now made its way into a respected scientific journal.<\/p>\n
Matteo Paz, a teen from Pasadena, didn\u2019t just stumble upon this cosmic treasure. Using artificial intelligence<\/a>, he opened a new window into the universe, showing how young minds can push the boundaries of space exploration.<\/p>\n Building an AI pipeline from scratchA university professor is convinced : pension payments to retirees must end<\/a>Jeff Bezos, founder of Amazon, in conflict with Elon Musk: \u201cWe can move heavy, polluting industries into space\u2014Earth should be reserved for living and light industry.\u201d<\/a><\/p>\n In the summer of 2022, Paz joined the Planet Finder Academy at Caltech, a program designed to expose high school students to real-world astronomy challenges. Under the mentorship of Davy Kirkpatrick from Caltech\u2019s Infrared Processing and Analysis Center (IPAC), Paz dove into NEOWISE, a NASA telescope archive.<\/p>\n Originally launched in 2009 to track near-Earth asteroids, NEOWISE ended up collecting over a decade of infrared data on the entire sky. This included not just nearby objects, but distant and often overlooked cosmic phenomena. The problem ? The archive contained nearly 200 billion lines of observations, far too massive for a small team to analyze manually.<\/p>\n Drawing on his background in theoretical mathematics, programming, and temporal analysis, Paz developed an AI model to automate the search. In just six weeks, he built a machine learning pipeline capable of detecting faint and variable light sources\u2014objects whose brightness changes over time and which typically escape human or conventional tools.<\/p>\n Discovering cosmic patterns<\/p>\n \u201cThe model showed promise almost immediately,\u201d Kirkpatrick told Phys.org. As Paz refined it, the results became increasingly fascinating. The AI identified objects that flickered, pulsed, or dimmed\u2014behaviors often linked to quasars, eclipsing binary stars, or supernovae.<\/p>\n What makes Paz\u2019s work remarkable is how it combines Big Data with deep space observation. By applying Fourier transforms and wavelet analysis\u2014mathematical techniques ideal for studying evolving signals\u2014the AI<\/a> detected subtle variations in the infrared spectrum that were previously invisible due to NEOWISE\u2019s temporal sampling limitations.<\/p>\n Some of these variable objects change so slowly or so briefly that they had gone unnoticed until now. This is critical for understanding slow transients and cataclysmic variables, which do not follow predictable patterns.<\/p>\n The impact on modern astronomyThe largest great white shark ever recorded spotted near this popular tourist destination<\/a>Everyone throws it away, but it contains 22-karat gold that could be worth thousands of dollars<\/a><\/p>\n Paz\u2019s achievement demonstrates that even decades-old data can reveal new cosmic insights when paired with innovative AI. It\u2019s a powerful reminder that breakthroughs don\u2019t always require the latest technology\u2014sometimes they just need fresh eyes and creative approaches.<\/p>\n For me personally, it recalls a lesson from high school : curiosity and persistence can turn ordinary projects into extraordinary discoveries. I remember staying up late analyzing simple datasets for a science fair project, thinking it was just a hobby. Who knew that persistence and creative thinking could lead to something groundbreaking ? Paz\u2019s work proves that the same principle applies on a cosmic scale.<\/p>\n Inspiring the next generation<\/p>\n