CoolFace
Apppublic

MohamedBiz/aura-lab-materials-discovery

sourceHugging Facemitupdated 8mo agoView on Hugging Face
0likes
App README

🚀 Aura Lab: AI-Powered Materials Discovery

Find the perfect materials for next-generation solid-state batteries!

What is this?

Aura Lab uses Physics-Informed Machine Learning (PIML) to predict the properties of battery materials in seconds instead of years. Our AI has analyzed 21,000+ materials and discovered 50 promising candidates for solid-state batteries.

Features

  • 🎯 Test a Material: Enter any lithium-based material formula and get instant predictions
  • 🏆 Our AI's Best Picks: See the top 10 materials ranked by overall performance
  • ℹ️ How It Works: Learn about our AI models and methodology

Predicted Properties

  1. 1.⚡ Speed (Ionic Conductivity): How fast lithium ions move through the material
  2. 2.🏗️ Strength (Stability): How stable and safe the material is
  3. 3.💎 Safety (Band Gap): How well the material prevents short circuits

Technology

  • Models: Random Forest, Gradient Boosting, XGBoost
  • Dataset: 21,307 lithium materials from Materials Project + 599 from OBELiX
  • Features: 41 engineered features (compositional, structural, electronic)
  • Performance: R²=0.670 (conductivity), R²=0.482 (stability), R²=0.778 (band gap)

Top Discovery

LiAlSiO4 (Eucryptite family)

  • Conductivity: 3.25×10⁻³ S/cm ⭐⭐⭐⭐
  • Stability: 0.0035 eV/atom ⭐⭐⭐⭐⭐
  • Band Gap: 5.06 eV ⭐⭐⭐⭐⭐
  • Overall Score: 0.870

About Aura Lab

We're building an AI-accelerated materials discovery platform to find novel solid-state battery electrolytes in 60 days instead of 10+ years. Join us in revolutionizing energy storage!

Contact: Learn More