Skip to content

markhryt/Devfest

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

26 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🚀 SkyHigh

An aerodynamic “frisbee robot” for rapid, low-cost data capture in rescue missions. SkyHigh glides silently, gathers images, audio, and environmental metrics, and delivers real-time insights via an interactive dashboard.


📦 Features

  • Silent Flight: Passive stabilization ensures low-noise deployment.
  • Data Capture: Camera, microphone, and environmental sensors record images, audio, temperature, and more.
  • Edge Filtering: RobotServer pre-selects high-quality data frames.
  • AI Analysis: AIServer leverages object detection, LLMs, and Groq.ai for actionable insights.
  • Real-Time Dashboard: Streamlit frontend visualizes maps, annotated frames, and live audio.

🛠️ Tech Stack

Hardware

  • Arduino & ESP32 (ESP-IDF)
  • Sensors: MPU-9250 (gyro, accelerometer, magnetometer), MEMS microphone, OV2640 camera
  • Passive Stabilization: Custom 3D-printed wings and bearings

Backend

  • FastAPI & Flask for RobotServer and AIServer
  • Groq.ai & OpenCV for AI inference
  • WebSockets for live data streaming

Frontend

  • Streamlit with custom HTML/Markdown components

🧩 Challenges

  • Mechanical Stability: Fine-tuning passive stabilization for consistent flight.
  • Streamlit Extensions: Embedding custom HTML/JS to overcome component limitations.
  • AI Agent Tuning: Balancing inference speed and accuracy on edge hardware.

🔮 What's Next

  • Aerodynamic Refinement: Test new wing profiles and materials for increased range.
  • Single-Battery Design: Improve power regulation to run on one cell.
  • Obstacle Avoidance: Develop simple steering mechanisms.
  • Post-Landing Mobility: Add wheels for ground traversal.
  • Launcher Mechanism: Create a spring or air-powered launcher for extended reach.
  • Boomerang Return: Experiment with curved designs for self-retrieval.

About

Devfest 2025 @ Columbia University

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C 59.7%
  • Python 22.6%
  • C++ 17.7%