Crop Advisory for Low-Internet Farmers (Offline-First)
Link to open source: https://github.com/HarshalGarade/hacknagpur.git
KrishiTech an mobile application
We propose an Offline-First Crop Advisory System designed specifically for low-internet farmers.
The system works without continuous internet access, provides easy-to-understand crop guidance, and automatically syncs data when connectivity becomes available.
The solution combines local storage, preloaded advisory content, and AI-based disease identification to ensure farmers receive timely, accurate, and practical advice anytime, anywhere.
Key Features of the Solution
1. Offline-First Architecture
The application works fully offline after initial installation.
All essential crop advisory data (crop details, growth stages, irrigation, fertilizer schedules, pest and disease information) is stored locally on the device.
Farmers can access information without internet, even in remote villages.
Benefits:
No dependency on mobile data
Reliable access during critical farming stages
Ideal for low-connectivity regions
2. Crop-Wise Advisory System
Farmers can select their crop from a predefined list (Rice, Wheat, Maize, Bajra, Jowar, Ragi, Cotton, Sugarcane, Vegetables, Fruits, etc.).
For each crop, the system provides:
Crop introduction
Soil and climate requirements
Growth stages
Water management
Fertilizer recommendations
Common pests and diseases
Preventive and corrective measures
All information is structured in simple steps suitable for farmers with limited digital literacy.
3. Offline Disease Identification (Image-Based)
Farmers can capture or upload a crop leaf image using their mobile device.
A pretrained plant disease detection model (PlantVillage) runs locally or semi-offline to identify:
Crop type
Disease type (if present)
The detected disease is matched with locally stored advisory data.
Output to Farmer:
Crop name
Disease name
Symptoms
Why the disease occurs
Preventive measures
Suggested fertilizer or treatment
This feature empowers farmers to take immediate action without waiting for experts.
4. Local Language Support (Optional but Recommended)
Advisory content is available in English, Hindi, and Marathi.
Farmers can select their preferred language during onboarding.
Language selection works offline.
Impact:
Improves understanding
Reduces dependency on text-heavy English apps
Increases adoption among rural users
5. Simple & Farmer-Friendly User Interface
Large buttons and icons
Minimal text with clear headings
Step-by-step flow (Select Crop → View Advice → Take Action)
Image-based navigation wherever possible
Designed keeping low digital literacy in mind.
6. Smart Sync Engine (When Internet is Available)
When internet connectivity returns:
App automatically syncs in the background
Updates advisory data (new diseases, updated recommendations)
Uploads anonymized usage data (optional)
Downloads improved AI model updates if available
No manual action required from the farmer.
7. Data Caching & Version Control
Advisory content is versioned
Only updated data is downloaded (low data usage)
Ensures farmer always has the latest verified guidance
8. Alert & Reminder System (Offline)
Reminds farmers about:
Irrigation schedules
Fertilizer application time
Disease-prone stages
Alerts work offline using local notifications
9. Expandable & Scalable Design
The system is designed to easily support:
New crops
New diseases
Regional farming practices
Weather-based advisories (when online)
Technical Architecture (High Level)
Frontend: Mobile App (Flutter / Android)
Backend: Lightweight API (Flask / FastAPI)
AI Model: Pretrained PlantVillage CNN
Storage: Local SQLite / JSON
Sync: Background sync using REST APIs
Languages: English, Hindi, Marathi
Expected Impact
Faster disease identification
Reduced crop losses
Better fertilizer usage
Improved farmer confidence
Increased yield and income
Strong digital inclusion for rural farmers
This build was uploaded as a hackathon project
