InfraBug

Inspiration

Cities rely on manual, slow, and opaque systems to report infrastructure defects. We wanted to modernize this process using AI while keeping privacy, trust, and accountability at the core.

What it does

InfraBug lets city workers capture photos of issues like potholes or cracked walls, get instant AI-powered defect analysis with repair recommendations, and generate visualization videos showing the repair process. All reports are privacy-safe and tamper-proof.

How we built it

We built a mobile-first application using on-device computer vision for image capture and redaction, Gemini AI for defect detection and repair insights, and blockchain to store immutable report metadata and verification proofs.

Challenges we ran into

• Ensuring privacy by blurring faces and license plates locally
• Designing reliable AI analysis for diverse infrastructure conditions
• Structuring blockchain data efficiently without high gas costs
• Keeping the app fast and usable for field workers

Accomplishments that we're proud of

• Fully automated, end-to-end defect reporting workflow
• Privacy-first design with local redaction
• Blockchain-backed accountability trail
• AI-generated repair visualization videos

What we learned

• Privacy and AI must be designed together, not added later
• Real-world infrastructure data is highly unstructured
• Blockchain works best when storing proofs, not raw data

What's next for InfraBug

• Expand defect categories and AI accuracy
• Add city-level dashboards and analytics
• Integrate predictive maintenance insights
• Pilot with real municipal bodies

Built With

• ether
• expo.io
• fast-api
• mongodb
• react-native
• render
• solidity