Next.js, React Three Fiber, FastAPI, scikit-learn, NASA/JPL SBDB
The problem
Judging an asteroid as a mining target means combining orbital data, likely composition and mission cost, which normally live in separate datasets and tools.
How it works
A FastAPI pipeline pulls live records from NASA/JPL's Small-Body Database and reclassifies them with K-Means clustering (k = 3).
Each target gets a briefing and a valuation from a theoretical economics model, clearly labelled as a simulation.
The dashboard ranks targets; selecting one loads its orbit in a React Three Fiber scene.
Admin pipeline runs are rate-limited and CORS is locked to known origins.
The result
934 near-Earth objects indexed and ranked, each with a mission briefing, an orbit view and a cost model.
Target acquisition table of 934 indexed near-Earth objects
Spectravein landing page: Live orbital intelligence
