🧑‍🌾 FarmSathi: Smart Agricultural Recommendation System

FarmSathi is an AI-powered agricultural assistant designed to support farmers in making informed decisions. The system provides comprehensive recommendations for crops, fertilizers, and crop disease management, ensuring better productivity and sustainable farming practices.

It integrates advanced machine learning models via a FastAPI backend and utilizes Firebase Authentication for secure user management.

✨ Core Features

FarmSathi integrates advanced machine learning models to deliver the following capabilities:

• Crop Recommendation: Suggests the best crops based on soil and environmental conditions (e.g., NPK, pH, Rainfall, Temperature).
• Fertilizer Recommendation: Recommends optimal fertilizers tailored to soil nutrient levels and crop type.
• Disease Detection: Identifies crop diseases from uploaded leaf images and suggests actionable remedies.
• Chatbot: Provides quick, context-aware answers to farming-related questions through an integrated widget.

---

💻 Frontend Overview

This repository contains the simple multi-page frontend for FarmSathi, built with HTML, CSS, and vanilla JavaScript (ES modules), and designed to interface seamlessly with the FastAPI backend.

Frontend File Structure

File	Description	Protection Status
index.html	Landing page with hero section and "Get Started" button.	Public
dashboard.html	Protected dashboard with feature cards.	Protected
crop-recommend.html	Form calling /crop-recommend.	Protected
fertilizer-recommend.html	Form calling /fertilizer-recommend.	Protected
disease-detect.html	Image upload for /predict-disease.	Protected
survey.html	Simple protected user survey form.	Protected
feedback.html	Simple protected feedback form.	Protected
contact.html	Simple protected contact form.	Protected
styles.css	Global styling and responsive layout.	N/A
firebase-config.js	Firebase configuration and API_BASE_URL definition.	N/A
app.js	Firebase initialization, Auth logic, route protection, and API helpers.	N/A
chatbot-widget.js	Floating chatbot button + modal calling /chatbot.	N/A

Authentication and Route Protection

• Firebase Authentication (Email/Password + Google Sign-In) is handled by app.js.
• All feature pages are protected.
• If a logged-out user tries to access a protected page, the auth modal opens. Upon successful login, the user is redirected back to the page they requested.

---

⚙️ How to Run Locally

1. Prerequisites

You must have:

1. A running FastAPI backend with the endpoints: /crop-recommend, /fertilizer-recommend, /predict-disease, and /chatbot.
2. Node.js or Python 3 to run a local HTTP server.

2. Configure Firebase and API URL

1. Firebase Setup: Go to the Firebase Console, create a Web App, and enable Email/Password and Google sign-in providers in Authentication settings.
2. Update Config: Open firebase-config.js and replace the placeholder keys with your Firebase configuration.
3. Set Backend URL: Verify the API_BASE_URL in firebase-config.js matches your backend's address (default: http://localhost:8000).

3. Start the Local Server

Navigate to the project root directory and start a local HTTP server.

❗ Note: Because the project uses ES Modules, opening index.html directly using a file:// path will cause errors. You must use an HTTP server.

Option	Command	Access URL
Option A (Node.js)	npx http-server .	http://localhost:8080
Option B (Python 3)	python -m http.server 8000	http://localhost:8000

---

🧪 Testing the Features

Once both the frontend and backend servers are running:

• Auth: Click Get Started and sign up/in. You should land on dashboard.html.
• Recommendation: Fill and submit forms on crop-recommend.html or fertilizer-recommend.html to receive predictions.
• Detection: Upload an image on disease-detect.html to get the predicted disease and recommendation.
• Chatbot: Click the floating chat bubble, ask a question, and verify the text response.
• Forms: Submit data on the survey/feedback/contact forms. Data is cached locally in localStorage for demo purposes.

---

🚧 Development Status

FarmSathi is currently under active development.

Implemented:

• Crop and fertilizer recommendation machine learning models (Backend).
• Complete Multi-page Frontend with Firebase Auth and API integration.

Upcoming Features:

• Integration of pesticide recommendation.
• Expansion of the knowledge base with pest and disease libraries.
• Scalability for additional features like irrigation suggestions.