Projects / NASA International Space Apps Challenge 2025

Award winnerGlobal ConnectionConfidence: high

Gaia+LEO

Optimization model that decides when and where orbital data centers beat terrestrial ones for AI and Earth-observation compute.

Original submission ↗View source ↗Watch video ↗Use as a reference →
Project facts
Track / challenge
Commercializing Low Earth Orbit (LEO)
Team
4 people · US
Build
software only
AI
None stated

Problem

Data-center growth for AI training and climate modeling keeps raising terrestrial power and water demand, and nobody can say when orbital compute would actually be cheaper or greener.

Solution

A mixed-integer optimization framework with an interactive Streamlit and Three.js interface that chooses which 40 MW solar-powered orbital compute modules to launch, which ground sites to activate and how to route workloads, minimizing capital and operating cost under latency, bandwidth, power and thermal limits, using NASA POWER and US energy-grid data.

Demo moment

Not described in the sources.

Why it stood out

  • organizer statedWon the award for the project that best connects people around the world through technology.Source ↗
  • organizer statedConnects the AI compute boom and the orbital economy with the goal of cutting terrestrial power and water demand.Source ↗
  • team statedQuantifies when and where orbital compute becomes advantageous instead of asserting it, with interactive scenarios for policymakers and cloud providers.Source ↗
  • curator inferenceOperations-research depth where most entries would draw a concept makes a speculative idea defensible.

Lessons for your next hackathon

  • Back a speculative concept with a real quantitative model and let judges move the sliders.
  • Tie the space idea to an Earth problem judges already feel (data-center power and water).

Similar situations

Projects that share domains, mechanisms or judging lenses — often with a different problem and stack.Tune the comparison.

Award winner · 2025

Astro Sweepers

Orbital debris compliance and risk intelligence: automated Debris Assessment Software runs, ready-to-file disposal reports and a risk index for every object.

Shared: domains space operations · mechanisms visible engineering depth · lenses impact, technical depth

42 QuakeHeroes — video preview▶ Video

Award winner · 2024

42 QuakeHeroes

A U-Net deep network on decomposed seismic energy that finds the onset of moonquakes and marsquakes in noisy lander data.

Shared: mechanisms visible engineering depth, trend wave timing · lenses technical depth · stack python

Podium #2 · 2024

BASIC Web

A from-scratch BASIC compiler and runtime for building interactive web pages.

Shared: mechanisms visible engineering depth · lenses technical depth, originality

Sources

  1. https://www.nasa.gov/learning-resources/stem-engagement-at-nasa/nasa-announces-2025-international-space-apps-challenge-global-winners/ · official results · retrieved 2026-09-14
  2. https://www.spaceappschallenge.org/2025/awards/ · official event page · retrieved 2026-09-14 — Award definition.
  3. https://www.spaceappschallenge.org/2025/find-a-team/gaialeo/?tab=project · project page · retrieved 2026-09-14 — Team-written summary, model, stack, AI-use disclosure and data.