Inspiration
Sports connect us all.
When we watch sports, we feel like part of something larger than ourselves. We hate the feeling of missing a big play while waiting in line for food and only hearing the sounds of cheers, but for the visually impaired, that’s their status quo. We wanted to enable them to enjoy sports as well by being more connected with what's going on.
What it does
- Tracks ball location and provides tactical experience that allows users to feel where the ball is relative to the field.
How we built it
- 3D printed and laser cut pieces designed from CAD
- Jetson Orin Nano that can receive json file data and use it to manipulate vibration sensors
- 9 relay boards to form a 5x4 grid
- OpenCV for visual tracking of the ball
An all-nighter
Challenges we ran into
- The ball tracking software uses color to identify the ball, however the player outfit was the same color as the ball so the software started tracking the player instead. also when the ball was moving extremely fast or when the ball was hidden (like football) it was hard to detect where it was
- It was hard to find videos where we could create a relational grid so we could map it to our field. Many sports cameras move a lot.
- 3D printing and laser cutting took an extremely long time
- Figuring out how to connect the 5x4 relay matrix was extremely difficult because we didn't fully understand the complexities of it.
Accomplishments that we're proud of
- figuring out how to make the 5x4 relay matrix grid on our own using fundamentals from linear algebra and circuitry, starting with an unclear idea of how to create it
- developing functional ball-tracking that you can visualize and can be used by the hardware parts
- having our 5x4 relay grid play marry had a little lamb
- having a complete project because at 2am we thought it wasn’t possible
What we learned
- It's very difficult to track a ball with software with normal sports videos. The ball can look like it's in the stands, or “disappear” into a player’s uniform
- Working with people we didn’t know and learning how to organize each person’s tasks based on their strengths
What's next for Buzzboard
- Retrieve data directly from sensors on the ball so it can provide more accurate, objective measurements of where the ball is relative to the field. This could work in any condition, not just with a nice overhead view.
- Add more vibration motors and find a way to increase resolution by using transistors and diodes.
- Add a button on the board that calls a Grok API to determine fan reactions to the game as well as game momentum, the importance of a certain drive, and use speech-to-text to inform the user so they can understand more about the game
Log in or sign up for Devpost to join the conversation.