SDG 6 · Clean Water and Sanitation | SDG 11 · Sustainable Cities | SDG 12 · Responsible Consumption
Problem Statement:
Many households, campuses, and municipal bodies lack awareness of their daily water consumption patterns. This leads to unintentional water wastage and undetected leakages, severely draining precious clean water resources and increasing costs.
Design Thinking Format: "How might we use AI to track and predict daily water usage so that households and campuses can become more sustainable?"
AI Solution:
AquaSense AI is a smart monitoring platform that tracks and analyzes water usage. It utilizes a Linear Regression model to predict future consumption trends, allowing users to plan ahead. Additionally, it features an Isolation Forest anomaly detection model that automatically flags unusual water usage spikes to identify potential hidden leakages early.
Target Users:
- Individuals and Households
- Campus Administrators and Facility Managers
- Local Municipal Bodies
Expected Impact:
If implemented on a large scale (e.g., across a college campus or a residential society), the solution will shift water management from a reactive approach (paying high bills after a leak) to a proactive approach (preventing waste before it happens). Consumers benefit economically, administrators can quickly dispatch maintenance, and the environment benefits from thousands of liters of conserved water.
Responsible AI Considerations:
- Fairness: The AI models rely purely on objective consumption metrics, completely avoiding demographic bias.
- Transparency: The dashboard clearly visualizes the data and provides understandable reasons when an anomaly is flagged.
- Ethics: Empowers users with awareness to promote conservation, rather than penalizing them.
- Privacy: User consumption logs and location data are kept strictly secure.
- Frontend – React 18 + Vite, Tailwind CSS, Recharts, Axios, React Router
- Backend – Node.js, Express.js, JWT Auth
- Database – MongoDB Atlas + Mongoose
- AI Service – Python Flask, Scikit-learn (Linear Regression + Isolation Forest)
# Backend
cd backend && npm install
# Frontend
cd ../frontend && npm install
# AI Service
cd ../ai-service && pip install -r requirements.txtCopy .env.example to .env in each directory and fill in your values.
backend/.env
MONGO_URI=mongodb+srv://<user>:<pass>@cluster.mongodb.net/aquasense
JWT_SECRET=your_super_secret_key
PORT=5000
AI_SERVICE_URL=http://localhost:5001
frontend/.env
VITE_API_URL=http://localhost:5000/api
ai-service/.env
PORT=5001
# Terminal 1 – Backend
cd backend && npm run dev
# Terminal 2 – Frontend
cd frontend && npm run dev
# Terminal 3 – AI Service
cd ai-service && python app.py
| Service | Platform |
|---|---|
| Frontend | Vercel |
| Backend | Render |
| AI Service | Render |
| Database | MongoDB Atlas |
cd frontend && npm run build
# Push to GitHub, connect repo in Vercel dashboard
# Set VITE_API_URL env var to your Render backend URL- Create two Web Services on Render
- Set root directories to
backend/andai-service/ - Add environment variables in Render dashboard
- Build command:
npm install/pip install -r requirements.txt - Start command:
node server.js/python app.py
- JWT Authentication (Register / Login / Protected Routes)
- Water consumption CRUD (date, liters, location, department, notes)
- Dashboard with live charts (daily/weekly/monthly trends)
- AI Predictions – Linear Regression (next day / week / month)
- Leakage Detection – Isolation Forest anomaly detection
- Sustainability Score (0–100) with level labels
- AI Recommendation Engine (dynamic, data-driven)
- PDF Report generation
- Admin Panel (all users, system analytics)
- Conservation Leaderboard
- Water Saving Goal Tracker
- CSV / Excel Import
- Email Alerts (Nodemailer)
- SDG-6 Impact Metrics page
aquasense-ai/
├── frontend/ React + Vite app
├── backend/ Express REST API
├── ai-service/ Flask ML microservice
└── docs/ Architecture diagrams