AI and IoT plant health platform combining ESP32 sensors with PyTorch disease detection.
Monitor crops with real-time ESP32 sensor streams, MobileNetV2 disease detection, and Gemma 4 AI chat.
Smart Plant Doctor is an open-source platform combining real-time environmental sensing from ESP32 devices with image-based disease classification and an AI care assistant. It suits agricultural engineers, smart farmers, and developers building custom IoT-driven crop monitoring systems.
The application ingests live sensor metrics like temperature, humidity, light, and soil moisture via WebSockets into a React dashboard. It classifies plant diseases across 29 classes using a PyTorch MobileNetV2 model and pairs diagnoses with Google Gemma 4 for contextual care advice.
It suits agtech developers and tech-savvy growers who want a full-stack system linking physical microcontrollers to AI-powered diagnostics. It does not suit non-technical users looking for a plug-and-play commercial product without manual hardware wiring and server configuration.
Streams temperature, humidity, light, and soil moisture from ESP32 devices via WebSockets.
Classifies plant diseases across 29 categories from uploaded leaf images using a trained PyTorch model.
Provides contextual care advice using Google AI integrated with live sensor and diagnosis data.
Tracks anonymized plant health reports using Leaflet and geohash data.
Free under its open-source MIT license. The only cost is running the required infrastructure, such as hosting the FastAPI backend, setting up a database, and covering API costs for Google Gemini.
880 open remote AI-training roles, updated November 15, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
These reviewed AI tools do the same job instantly — from our scored directory.
You need an ESP32 microcontroller configured with the provided PlatformIO firmware to ingest temperature, humidity, and soil moisture data.
The MobileNetV2-based model reports a validation accuracy of approximately 92.37% across 29 classes for six supported plant species.
Yes, local development supports SQLite alongside PostgreSQL, though Supabase is required for Google OAuth and managed cloud auth.
See Zekai first in Google