im-amrith/crisp

CrispRoot Platform 🌾

A comprehensive agricultural technology platform that transforms traditional farming into a smart, digital-first experience. This system serves as a real-time digital twin of farms, empowering farmers, agronomists, and developers with AI-driven insights, predictive analytics, and intelligent monitoring capabilities

Overview

It reimagines agriculture through four core domains: market analysis, crop simulation, plant disease detection, and livestock health management. The platform combines data-driven decision making with real-time monitoring to optimize agricultural operations, increase profitability, and ensure sustainable farming practices.

✨ Key Features

🧠 AI Garden Advisor

• —Personalized Recommendations: Tailored crop suggestions using Google Gemini AI.
• —Multi-language Support: Translations available in 10+ Indian languages.
• —Smart Input Processing: Considers location, soil type, budget, and experience level.

📊 Market Demand Analysis

• —Predictive Analytics: Neural Perceptron model for 12-month demand forecasting.
• —Market Intelligence: Seasonality analysis and optimal selling period recommendations.
• —Pricing Strategy: Net market price calculations including transportation costs.
• —Data Source: Variety-wise Daily Market Prices from Open Government Data Platform.

🌱 Crop Simulation & Growth Monitoring

• —PCSE Framework: Python Crop Simulation Environment based on WOFOST principles.
• —Real-time Monitoring: Daily plant growth simulation with live updates.
• —Growth Metrics: Yield predictions, biomass accumulation, Leaf Area Index (LAI).
• —BBCH Integration: Growth stage identification based on research standards.

🔍 Plant Disease Detection

• —High Accuracy: 90% accuracy using MobileNetV2 architecture.
• —Comprehensive Coverage: Detects 80 different plant diseases.
• —Dataset: Trained on Plant Village and 20K Crop Disease datasets.
• —Instant Results: Real-time disease identification with confidence scores.

🐄 Livestock Disease Prediction

Dual-Model Approach:

• —Symptom-Based: 80% accuracy using Random Forest algorithm for 5 major diseases.
• —Appearance-Based: 70% accuracy using MobileNetV2 for 7 cattle conditions.
• —Disease Coverage: Pneumonia, Lumpy Skin Virus, Blackleg, Foot and Mouth Disease, Anthrax.

🌍 Green Guardian

• —Environmental Monitoring: Real-time weather alerts and emergency response.
• —Interactive Features: Chat interface, live dashboard, map views.
• —Satellite Integration: Satellite imagery analysis for comprehensive monitoring.

🎤 Multilingual AI Assistant

• —Voice & Text Support: Available in 10+ Indian languages.
• —Call-based AI Bot: Phone support for farmers with feature phones.
• —WhatsApp Integration: Bot support for broader accessibility.

🌾 Crop Management Advisor

• —Contextual Crop Planning: Recommends suitable crops based on prior crop history, pest issues (e.g., rat attacks), maintenance levels, and land conditions.
• —Goal-Based Suggestions: Adapts to the farmer’s goals—like high yield, low maintenance, or short-duration harvests.
• —Smart Estimations: Provides estimated yield, growth duration, and profitability metrics for each crop option.
• —Dynamic Adjustment: Learns from outcomes and feedback to refine future recommendations.

🌿 Biosphere & Integrated Farming Modeling

• —Multi-layer Farming Simulation: Supports co-cultivation systems like fish-rice farming.
• —Resource Optimization: Enhances productivity and land usage through biodiversity-driven planning.
• —Sustainability Focused: Promotes climate-resilient and regenerative agricultural practices.

🧬 CRISP Gene Compatibility Scoring

• —Hybridization Predictor: Evaluates the feasibility of cross-breeding two plant varieties.
• —Data-Driven Scoring: Uses genomic and phenotypic data to assign a compatibility score.
• —Breeder Guidance: Helps select optimal genetic combinations for higher yield, disease resistance, and adaptability.

🧩 CrispRoot Interaction Architecture

[image]

“Farmer-Centric Smart Agriculture Architecture: CrispRoot Platform”

This architecture diagram represents how a farmer interacts with the CrispRoot platform, flowing through a multi-layered intelligent system that bridges user actions with data-driven insights and automation. The platform serves as a digital twin of a farm, offering real-time decision-making, monitoring, and AI-powered advisory across agriculture domains.

🔹 1. Farmer Interaction Layer

Label: Farmer (User)

• —Access via smartphone, tablet, or computer
• —Interaction Types:
• —Voice command
• —Image upload (leaf/livestock)
• —Manual input (text/forms)
• —Multilingual UI (10+ Indian languages)
• —Offline support via call-based AI bot

🔹 2. Presentation Layer

• —Mobile App – Built with React Native
• —Web Portal – Built with React.js
• —PWA Features:
• —Offline sync
• —Push notifications
• —Responsive layouts for all devices

🔹 3. API Gateway Layer

• —Unified API Gateway handling:
• —OAuth 2.0 & Multi-Factor Authentication
• —Biometric login
• —Secure routing & rate limiting

🔹 4. Core Microservices (Business Logic)

🔹 5. Data Processing & AI Layer

• —Streaming: Apache Kafka
• —Batch Processing: Apache Spark
• —MLOps Capabilities:
• —Automated model retraining
• —A/B testing
• —Version control
• —Performance monitoring

AI Models in Use:

• —MobileNetV2 (Plant & Livestock Disease)
• —Neural Perceptron (Market Demand Forecast)
• —PCSE/WOFOST (Crop Simulation)
• —Random Forest (Livestock Symptom Analysis)

🔹 6. Data Layer – Polyglot Storage

• —Firebase Firestore – Real-time farm data
• —PostgreSQL – Structured crop & user data
• —MongoDB – Unstructured logs & media
• —Redis – Fast caching for performance
• —Cloud Storage (AWS S3) – Leaf/livestock images

🔹 7. External APIs & Data Feeds

• —Weather – OpenWeatherMap API
• —Soil – ISRIC SoilGrids API
• —Market – Agmarknet API
• —Govt. Schemes – India Open Government API

🔹 8. Feedback Loop (Learning Cycle)

This loop ensures adaptive intelligence where the system improves over time using real-world feedback and seasonal variation.

🛠️ Tech Stack

Frontend

• —React with TypeScript
• —Next.js for server-side rendering
• —Vite for fast development builds
• —Tailwind CSS + PostCSS for styling
• —Chart.js & Recharts for data visualization
• —i18next for internationalization

Backend & Services

• —Node.js with TypeScript
• —PostgreSQL database
• —Prisma ORM for database interactions
• —tRPC for type-safe API layer
• —Python Flask for ML model inference
• —Google Gemini AI for recommendations and translations
• —Firebase for cattle farm management

Machine Learning

• —MobileNetV2 for disease detection
• —Random Forest for livestock symptom analysis
• —Neural Perceptron for market demand forecasting
• —PCSE (Python Crop Simulation Environment)

🚀 How to Run

Prerequisites

• —Node.js (v18 or higher)
• —Python 3.8+
• —PostgreSQL database
• —Firebase account
• —Google Gemini API key

Installation

1. 1.Clone the repository

git clone https://github.com/annam-ai-iitropar/team_1B.git

1. 1.Install dependencies

# Frontend dependencies
npm install

# Backend dependencies
cd backend
npm install

# Python ML service
cd ../ml-service
pip install -r requirements.txt

1. 1.Environment Setup

# Copy environment files
cp .env.example .env.local
cp backend/.env.example backend/.env
cp ml-service/.env.example ml-service/.env

1. 1.Database Setup

# Run Prisma migrations
cd backend
npx prisma migrate dev
npx prisma generate

1. 1.Start the services

# Start ML service
cd ml-service
python app.py

# Start backend
cd ../backend
npm run dev

# Start frontend
cd ../
npm run dev

Configuration

Update the following environment variables:

• —DATABASE_URL: PostgreSQL connection string
• —GOOGLE_GEMINI_API_KEY: Google Gemini API key
• —FIREBASE_CONFIG: Firebase configuration
• —WEATHER_API_KEY: Weather service API key

📱 Accessibility Features

• —Responsive Design: Works on smartphones, tablets, and desktop
• —Voice Integration: Voice-to-text and text-to-speech capabilities
• —Multi-language Support: 10+ Indian regional languages
• —Offline Capability: Phone call support for feature phones
• —Gamification: Engaging user experience with gamified elements

🔄 System Architecture

The platform follows a modular architecture with clean separation of domains:

1. 1.Data Layer: PostgreSQL database stores farms, plots, crops, and sensor data
2. 2.API Layer: tRPC provides type-safe communication between frontend and backend
3. 3.ML Layer: Python Flask service hosts machine learning models
4. 4.Frontend Layer: React-based dashboard with real-time updates
5. 5.AI Layer: Google Gemini AI for recommendations and translations

📊 Performance Metrics

• —Plant Disease Detection: 90% accuracy
• —Livestock Symptom Analysis: 80% accuracy
• —Livestock Appearance Analysis: 70% accuracy
• —Market Prediction: 12-month forecasting capability
• —Real-time Updates: 2-30 second configurable refresh intervals

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• —Plant Village Dataset and 20K Crop Disease Dataset from Kaggle.
• —BBCH Scale Documentation for crop growth modeling.
• —Open Government Data Platform for market price data.
• —SLIIT-KUEMD Dataset for cattle disease prediction.

Built with ❤️ for the farming community