Inspiration
Rather than drawing inspiration from specific companies, our idea emerged from a vision of harnessing AI to make traffic and weather monitoring smarter and safer for people.
What it does
VisionFlow AI processes images from Seattle every five minutes, using advanced AI to assess traffic conditions, detect car accidents, and analyze weather patterns. The system ranks congestion levels and identifies accidents with binary classifications (0 or 1), while providing insights into current weather conditions based on the visual data.
How we built it
We developed the AI model using Python for its robust machine-learning capabilities, while the front end was built using HTML and React to ensure a smooth and user-friendly experience for our platform.
Challenges we ran into
One significant challenge arose when a team member lost access to our EC2 instance for over an hour, which temporarily delayed progress on their portion of the project. Despite this setback, we collaborated efficiently to ensure the project stayed on track. Another challenge we went through was trying to get our map to work which took over 2 hours to get it to finally work on the website while it took in excess of 6 more hours to get the data to work with it.
Accomplishments that we're proud of
We successfully deployed our application on an EC2 instance, integrated a MySQL database using RDS, and utilized Anthropic Claude 3.5 Sonnet to analyze INRIX traffic images for detecting car accidents, congestion, and weather conditions. Achieving this level of functionality and integration was a major milestone for our team.
What we learned;
We gained valuable experience in setting up a web server on AWS and collaborating effectively as a team to build a cohesive application. Additionally, we learned how to harness real-world data and AI models to create a system that accelerates accident detection and addresses traffic challenges, providing practical solutions with real-world impact.
What's next for VisionFlow AI
The future of VisionFlow AI includes integrating our interactive map while expanding access to live camera feeds nationwide and globally, and empowering users everywhere with real-time traffic and weather updates. Additionally, our crash detection feature could be enhanced to automatically notify emergency services, potentially saving lives by streamlining the reporting process.
Log in or sign up for Devpost to join the conversation.