🇺🇸Amber
Yang

There are approximately 500,000 pieces of debris larger than a marble orbiting Earth at speeds exceeding 17,000 miles per hour. Each one is a potential collision threat to the satellites that underpin GPS navigation, weather forecasting, telecommunications, and military surveillance. When two objects collide in orbit, they generate hundreds of additional fragments, each of which can generate more. Physicists call the worst-case scenario Kessler Syndrome — a cascade of collisions that could render large portions of low Earth orbit permanently unusable for human civilization. Amber Yang, seventeen years old, decided to do something about it.

Yang built a machine learning system capable of predicting the future trajectories of space debris with greater accuracy than existing methods. The core challenge is computational: tracking hundreds of thousands of objects requires modeling the gravitational effects of Earth, the Moon, the Sun, atmospheric drag at extreme altitudes, and solar radiation pressure — a system of equations that becomes intractable quickly. Existing tracking systems relied on numerical integration methods that computed each object's path step by step, an approach that scaled poorly with the size of the debris population. Yang's approach trained a neural network to learn the underlying dynamics directly from historical trajectory data, producing a model that could extrapolate future positions far more efficiently.

The project was called Seer — a name that captured the predictive ambition of the system. Yang had no formal training in orbital mechanics or machine learning beyond what she taught herself. She read research papers, worked through the mathematics of gravitational physics, and built the model over months of independent work. When she presented Seer at the Intel International Science and Engineering Fair in 2017, judges representing the scientific community were sufficiently impressed that she was awarded the Gordon E. Moore Award — the fair's highest honor, carrying a $75,000 prize — and named one of the grand winners of a competition that draws nearly 1,800 finalists from seventy-five countries.

The Intel ISEF is the largest pre-college science competition in the world. Winning a grand prize there — as opposed to a category award — is the kind of recognition that signals work at the frontier of a field, not just impressive schoolwork. Yang's system addressed an active area of concern at NASA, the European Space Agency, and private launch companies including SpaceX and RocketLab. The commercial space industry was growing rapidly in 2017, with launch frequencies rising and the debris population expanding in parallel. A faster, more accurate debris-tracking system had immediate practical applications.

Yang subsequently enrolled at Stanford University to study physics and computer science — a combination that positioned her at the intersection of the theoretical and applied research needed to push the space debris problem toward industrial-scale solutions. Her work was cited in discussions of space situational awareness, the technical term for the ability of spacefaring nations and companies to know where things are in orbit and predict where they will be. In a domain where a collision can happen in milliseconds and the debris can spread across thousands of kilometers of orbital shell, situational awareness is the difference between the sustainable use of space and a catastrophic cascade.

Yang's story sits at the intersection of two accelerating curves: the explosion of machine learning capability from 2012 onward, and the rapid expansion of commercial spaceflight following SpaceX's Falcon 9 reusability demonstrations in 2015 and 2016. She saw before most adults in the field that the tools of the AI revolution were applicable to the physical problems of the space environment, and she built a working system to prove it — at seventeen, before finishing high school, with no institutional support beyond her own curiosity and the availability of research literature and open-source machine learning frameworks. The orbital environment she is helping to protect is the infrastructure layer that the modern world runs on.