RoomScan

Turn room photos into an AI-powered appliance inventory and instant electricity bill estimate.

Comment

Inspiration

Managing household electricity consumption is difficult because most people don't know exactly how much energy each appliance uses. Creating an appliance inventory manually is tedious, and estimating electricity bills often involves guesswork. We wanted to build an AI-powered solution that makes this process effortless. By leveraging Google's Gemini multimodal capabilities, we created RoomScan—a tool that can analyze room photos, identify electrical appliances, and provide an estimated monthly electricity bill in just a few clicks.

What it does

RoomScan allows users to upload up to three photos of a room, where Gemini AI automatically detects visible electrical appliances and estimates their power ratings. The detected appliances are presented in an editable inventory, allowing users to adjust wattage, quantity, and daily usage hours before calculating energy consumption and estimated monthly electricity costs. Users can also export the inventory as a CSV file and customize electricity tariffs based on their region.

Key features include:

  • 📷 AI-powered appliance detection from room images
  • ⚡ Automatic wattage estimation
  • ✏️ Editable appliance inventory
  • 💰 Monthly electricity bill estimation
  • 📊 Appliance-wise energy consumption breakdown
  • 📁 CSV export
  • 💾 Local auto-save for a seamless user experience

How we built it

RoomScan was built using a modern full-stack architecture.

Frontend

  • Next.js 14 (App Router)
  • React
  • TypeScript
  • Responsive UI
  • Local Storage for persistence

Backend

  • Next.js API Routes
  • Google Gemini Vision API
  • Server-side image preprocessing
  • Request validation and rate limiting

Workflow

  1. Users upload room images.
  2. Images are optimized and securely sent to Gemini Vision.
  3. Gemini identifies electrical appliances and estimates power ratings.
  4. The application generates an editable inventory.
  5. Users customize appliance usage and electricity tariffs.
  6. RoomScan calculates monthly energy consumption and estimated electricity costs.

Challenges we ran into

One of our biggest challenges was ensuring reliable appliance detection across different room layouts, lighting conditions, and camera angles. Some appliances were partially visible or visually similar, requiring careful prompt engineering and image preprocessing to improve Gemini's accuracy.

Another challenge was balancing performance and API costs. High-resolution images increased processing time, so we implemented server-side image optimization before sending requests to Gemini.

We also focused on security by ensuring the Gemini API key remained on the server while implementing validation and rate limiting to prevent misuse.

Accomplishments that we're proud of

  • Successfully integrated Gemini Vision to perform real-world appliance detection.
  • Built an end-to-end workflow from room photos to electricity bill estimation.
  • Created an intuitive interface where AI-generated results can be reviewed and edited instead of blindly accepted.
  • Added useful features such as CSV export, customizable electricity tariffs, and automatic progress saving.
  • Developed a practical AI application that demonstrates how multimodal models can solve everyday problems.

What we learned

This project gave us hands-on experience with multimodal AI, prompt engineering, image optimization, and secure API integration. We learned that AI-generated predictions become far more valuable when users can verify and edit the results. We also improved our understanding of building scalable Next.js applications, handling image processing pipelines, and creating intuitive user experiences around AI-generated data.

What's next for RoomScan

We plan to expand RoomScan into a complete AI-powered home energy assistant by adding:

  • 🏠 Multi-room and whole-home scanning
  • 📈 Historical energy usage analytics
  • ☁️ Cloud synchronization across devices
  • 📱 Progressive Web App (PWA) support
  • 🔌 Smart meter and IoT integration
  • 🌍 Support for more regional electricity tariffs
  • 🤖 Personalized energy-saving recommendations powered by Gemini
  • 📷 Improved appliance recognition with higher accuracy and confidence scoring

Our long-term vision is to help homeowners better understand their energy consumption, reduce electricity costs, and make smarter, more sustainable decisions through AI.

Built With

Share this project:

Updates