PlanetAI: The Interactive Cosmic Explorer
Inspiration
PlanetAI was born from a simple question: How can we make space education as vast and awe-inspiring as space itself? Traditional textbooks often fail to capture the scale and dynamic nature of our universe. We wanted to build a bridge between complex astronomical data and students, creating a "living" star-map that feels more like a voyage than a lesson.
What it does
PlanetAI is an all-in-one educational 3D platform that allows users to:
- Explore a 3D Solar System: Interactive Three.js model with realistic textures, orbit paths, and cinematic camera controls.
- Interact with NASA Live Data: Fetches real-time imagery from the DSCOVR satellite (EPIC), latest Mars weather from the InSight lander, and tracks Near-Earth Objects (asteroids) approaching Earth today.
- Learn with AI: A built-in "Nova" Assistant (powered by Groq) that answers complex space questions and generates dynamic "Galactic Quizzes" to test user knowledge on-the-fly.
- Experience Premium UX: A futuristic interface featuring glassmorphism, Lottie animations, and a responsive design tailored for the modern explorer.
How we built it
The project was built using a "no-framework" approach to maximize performance and control:
- Core: Semantic HTML5, CSS3 (Custom Design System), and Vanilla JavaScript (ES6+).
- 3D Engine: Three.js for rendering the solar system, utilizing
OrbitControls,Raycasterfor selection, andSpriteMaterialfor dynamic labels. - APIs:
- NASA Open APIs: APOD, EPIC, Mars Photos, NEO, and InSight Weather.
- Groq Cloud API: Utilizing the
mixtral-8x7b-32768model for ultra-fast, intelligent conversational responses and quiz generation.
- Animations: Lottie for high-fidelity vector animations and CSS IntersectionObserver for scroll-triggered entrance effects.
Challenges we ran into
- 3D Scaling: Representing the solar system to scale while keeping it navigable was a challenge. We implemented a hybrid scaling system to ensure the Sun didn't dwarf the terrestrial planets while maintaining visual clarity.
- Async Data Orchestration: Fetching data from multiple NASA endpoints simultaneously could cause UI lag. We implemented robust error handling and loading states (Warp Drive Loader) to ensure a smooth user experience.
- AI Formatting: Getting the AI to consistently output structured JSON for the Quiz Master required precise system prompting and post-processing logic.
Accomplishments that we're proud of
- Ultra-Fast Performance: By avoiding heavy frameworks, we achieved a near-instant load time even with a complex 3D scene.
- Dynamic AI Integration: Creating a quiz system that never repeats the same question because it's generated by AI in real-time.
- Visual Fidelity: Achieving a "AAA Game" aesthetic using only standard web technologies and glassmorphism.
What we learned
- Three.js Optimization: Learned the importance of texture mapping and lighting to create realistic planetary bodies.
- Prompt Engineering: Refined our ability to guide LLMs to provide educational, safe, and structured responses.
- UX for Data: Figured out how to present "scary" scientific data (like asteroid proximity) in a way that is engaging and informative rather than overwhelming.
What's next for PlanetAI
- VR/AR Integration: Bringing the solar system into WebXR for a truly immersive classroom experience.
- Multiplayer Mode: Allowing students to join a shared "Space Lobby" and explore the cosmos together.
- Advanced Physics: Adding gravity simulations and "What-If" scenarios (e.g., "What if Jupiter was as close as the Moon?").

Log in or sign up for Devpost to join the conversation.