FindAir

Snap, tag, notify. Lost items returned efficiently.

Comment

🚀 Inspiration

Losing items on flights suck, and the current lost and found process lacks tracking and communication. We aimed to transform item recovery at American Airlines with a seamless application that reassures passengers and integrates effortlessly into airline operations—ensuring minimal impact on flight schedules.

🎉 What It Does

FindAir is an innovative lost item tracking system designed for airline cleaning crews to efficiently record and notify passengers about misplaced items. The workflow includes:

  1. Flight Verification: Employees enter the flight number and verify details with the American Airlines API.
  2. Multimodal Input System: Employees use their phone to:
    • Scan the item using their camera.
    • Recite the seat number via a privacy-preserving mic toggle.
  3. AI-Powered Processing:
    • Images are uploaded to Cloudinary.
    • LLaMA 3.3 (powered by Groq) generates item descriptions.
    • Seat and flight details are cross-referenced to retrieve passenger contact information.
  4. Automated Notifications:
    • Emails are sent to passengers with item details, retrieval QR codes, and shipping options.
    • Nearby passengers also receive alerts in case of shifted items.
  5. Claim Verification:
    • Employees at the lost and found booth can scan QR codes to mark items as claimed in the system.

🛠️ How We Built It

  • Frontend: Developed with Next.js for an intuitive user interface and hosted on Vercel for scalability.
  • Backend: Integrated with the American Airlines API for flight verification and data retrieval.
  • Image Hosting: Utilized Cloudinary for efficient image storage and processing.
  • AI Processing: Leveraged Groq's LLaMA 3.3 for rapid and accurate item descriptions.
  • Database: Stored seat-passenger mapping data to facilitate quick lookups and notifications.

Blendr-Page-1-5.png

💧 Challenges We Ran Into

  • Ensuring Speed and Efficiency: Balancing the need for accuracy with real-time processing to keep flights on schedule.
  • AI Description Accuracy: Fine-tuning LLaMA 3.3 to provide meaningful and accurate descriptions of diverse items.
  • User Experience: Designing a workflow intuitive enough for cleaning crew members to adopt without extensive training.
  • Data Privacy: Ensuring passenger data is handled securely only handling necessary information and Role based Access Control

🏆 Accomplishments That We're Proud Of

  • Successfully implemented multimodal input (voice and image) for faster item ingestion.
  • Automated email notifications that significantly improve the passenger experience.
  • A potential solution to increase item recovery rates beyond the 30-40% benchmark seen at major airports like London Gatwick.

📚 What We Learned

  • The critical role of intuitive UI design in adoption by non-technical airline staff.
  • Optimizing cloud-based AI workflows to ensure real-time responsiveness.
  • The importance of flight punctuality and how technology solutions must align with operational efficiency.

🔮 What's Next for FindAir

  • Expansion to Other Airlines: Collaborate with other major airlines to implement FindAir at a larger scale.
  • Enhanced AI Models: Improve item recognition accuracy and incorporate multilingual support for a global passenger base.
  • Mobile App Integration: Develop an app for passengers to track and report lost items directly.
  • Shipping Partnerships: Work with logistics companies to streamline item delivery to passengers' homes.

💻 Tech Stack

  • Frontend: Next.js, TailwindCSS
  • Backend: Node.js, Vercel
  • AI Model: Groq (LLaMA 3.3)
  • Image Hosting: Cloudinary
  • Database: Airline API Integration
  • Deployment: Vercel

By leveraging modern AI technologies and thoughtful system design, FindAir aims to revolutionize how airlines manage lost items, making the process seamless, efficient, and passenger-friendly.

Built With

Share this project:

Updates