StructSure

Leveraging AI to assess bridge damage.

Comment

Inspiration

The collapse of the Francis Scott Key Bridge struck a chord with us, highlighting the importance for efficient analysis tools in disaster responses. It inspired us to leverage the hackathon platform to contribute effectively to the bridge's repair efforts, aiming to assist in rapid recovery and safety.

What it does

StructSure is designed to analyze images of bridges, to identify zones of critical damage and zones of least critical damage.

How we built it

We developed "StructSure" using a stack of current technologies and services. The core functionalities are powered by Microsoft Azure, particularly Azure Blob Storage for managing image data and Azure Custom Vision for analyzing and assessing damage through AI-driven image recognition. The application is built on Next.js and React, leveraging TypeScript for robust backend operations managed via Node.js. For the frontend, TailwindCSS ensures a responsive and intuitive UI. The entire application is hosted and deployed through Vercel, offering a seamless, scalable cloud solution.

Challenges we ran into

Training the AI model with Azure Custom Vision and integrating the Custom Vision into our front were the 2 most challenging parts of the project.

Accomplishments that we're proud of

Despite being newcomers to AI and ML, we successfully created a functional web-app capable of performing image analysis to aid in disaster-recovery.

What we learned

This project was a profound learning experience. We got to understand how machine learning training works with images and learning new technologies in a full-stack setting to help apply our solution to a real-world situation.

What's next for StructSure

We plan to add more features to the Web App to give helpful insights for construction workers in safely clearing collapsed bridges. Additionally, we'd like to feed the model more data to better survey areas to warn construction workers.

Built With

Share this project:

Updates