Possible Jam

Traffic track of AWSxInrix Hackathon. The API system to use to analyze traffic intensity given an image of traffic.

Comment

Instructions

To try out this project:

  1. Clone the repository (git clone https://github.com/pyoung2/AWSxInrix-Hackathon.git).
  2. Change directory into the backend directory of the repository (cd AWSxInrix-Hackathon/backend).
  3. Run python api.py. May have to: pip install bedrock, flask, flask_cors, boto3, dotenv.
  4. Navigate to: -http://127.0.0.1:5000/api -http://127.0.0.1:5000/api/images -http://127.0.0.1:5000/api/images?image={IMAGE_NAME} -http://127.0.0.1:5000/api/bedrock?image={IMAGE_NAME}
  5. Poke around!

Inspiration

We took inspiration from last year's 2nd-place transportation track winner, called Littercator.

What it does

Possible Jam determines the intensity of traffic based on an image of cars. It gives a detailed analysis of the traffic and rates the traffic from a scale of 0 (no traffic) to 100 (high traffic). Quickly identifying traffic intensity can help with traffic control. Alongside the timestamps attached to each image, patterns may emerge to indicate time periods of low or high traffic intensity.

How we built it

We used Python, mainly Bedrock and Flask, for the backend. The Flask API referenced the Bedrock Claude AI model API. We were planning to use React alongside JS, HTML, and CSS for the frontend, and ended up using HTML in Flask.

Challenges we ran into

We struggled with creating the frontend due to a lack of experience and skill with website development, and "programming languages" like JavaScript, HTML, and CSS.

Accomplishments that we're proud of

Programming a working API system with integrated artificial intelligence tailored to the goal of analyzing traffic.

What we learned

We learned how to create an API, how to use Flask to make an API server, and how to use a Bedrock AI API.

What's next for Possible Jam

Implementing Google Maps into the frontend. Hosting the website and API server on AWS via EC2 will enable easy public access without the need to fork the GitHub repository and run it locally. Use the Inrix API to get more information about the traffic images for more features.

Share this project:

Updates