Inspiration
In Indian homes, lights and fans stay ON in empty rooms for hours — my parents' electricity bill was ₹3,200/month, 30% just waste. We realized: people don't waste power on purpose, they just forget. We needed a system that sees waste in real-time and shouts "band karo!"
What it does
SmartEnergySaver AI is a live dashboard that:
- Detects motion + light levels in each room (PIR + LDR sensors)
- Flags waste in 60 seconds: "Living Room light ON, no motion for 23 min"
- Shows exact money loss: "₹0.23 wasted today"
- One-tap OFF from phone via Firebase
- 7-day analytics to track savings
Live: https://parineetapahade86-maker.github.io/smartenergysaver-ai/
How we built it
- Frontend: Vanilla JS + Chart.js (no frameworks, loads in 1.2s)
- Backend: Firebase Realtime Database for instant sync
- Hardware (prototype): ESP32 + PIR, LDR, DHT22, ACS712
- Design: Mobile-first glassmorphism UI, works on ₹7k phones
We built it as a single HTML file so anyone can run it — just open index.html.
Challenges we ran into
- Firebase free tier disconnects — solved with auto-reconnect + local demo mode
- Making waste feel real — added ₹ counter instead of just kWh
- GitHub Pages doesn't support backend — converted everything to client-side with Firebase
- Time crunch — built core in 6 hours, polished UI overnight
Accomplishments that we're proud of
- Working live demo in <24 hours
- Real ₹ savings calculation (not fake data)
- Deployed on both Netlify AND GitHub Pages
- Mobile responsive, works offline in demo mode
- Zero-cost stack (Firebase free + GitHub Pages)
What we learned
- Real-time UX > fancy AI — users want instant feedback
- Simple math (watts × time × ₹ rate) beats complex ML for this problem
- Single-file deployment wins hackathons
What's next for SmartEnergySaver AI
- Auto-cutoff via smart plugs (Tuya API)
- WhatsApp alerts: "Bedroom fan ON since 2 hours"
- College hostel pilot — target 20% bill reduction
- Hindi voice alerts for parents

Log in or sign up for Devpost to join the conversation.