A hardware-agnostic, fail-proof platform for autonomous hydroponics & aeroponics. Powered by Zero-Shot AI (SCOLD) and Google Gemini.
Yield is a modular operating system for smart agriculture. Unlike traditional systems that lock you into specific vendors, Yield is designed to be hardware-agnostic --- working seamlessly with ESP32, Arduino, Raspberry Pi, or industrial PLCs.
It solves the critical "Fail-Point" problem in aeroponics (pump failures/sensor drift) through a redundant watchdog architecture and employs state-of-the-art Zero-Shot Computer Vision to detect plant diseases without needing custom training data.
- 🧠 Zero-Shot Disease Detection: Powered by the SCOLD model (CLIP + RoBERTa).
- 🤖 AI Agronomist: Integrated Google Gemini API analyzes sensor logs.
- 📡 Universal Connectivity: HTTP-based IoT Gateway supporting ESP32 & ESP32-CAM.
- 🛡️ Fail-Proof Architecture: Real-time heartbeats and fallback mechanisms.
- 👥 Role-Based Interface:
- User Panel: Monitor growth and alerts.
- Owner Panel: Manual control over nozzles, lights, and vision queries.
Place all images inside a folder named
screenshotsin your project root.
The system operates as a cohesive Monorepo with three distinct layers:
- The Brain (Backend): FastAPI orchestrator handling logic and MongoDB.
- The Eyes (SCOLD Engine): PyTorch service for Vision Transformers.
- The Face (Frontend): React + Vite application for visualization.
- Python 3.9+
- Node.js 18+
- MongoDB (Local or Atlas URL)
Ubaid Patel
Built with ❤️ for the future of Sustainable Agriculture.