ChessMate

• 

  Logo

• 

  Team

• 

  Computer Vision

• 

  Contraption

• 

  Chessboard

• 

  Group Photo

• 

  Wiring

• 

  Cutting

• 

  Stand

🌍 Inspiration

• All of us come from immigrant families and wanted to relive the experiences of playing chess with family abroad.
• ChessMate brings the joy of real chess to people separated by distance. Where one person plays virtually, and the pieces move on the board in real-time.
• We also wanted to make chess more inclusive and accessible, especially for those who may struggle with mobility or vision.

♟️ What it does

• ChessMate is a smart chess companion and assistant that helps you improve at chess, whether playing alone or virtually with others
• It blends physical and digital gameplay by moving pieces on a real chessboard through a robotic gantry system.
• Features accessibility options such as voice commands to move pieces and AI guidance for coaching and practice.

🛠️ How we built it

• ChessMate is powered by a 2D, dual-axis gantry that perceives, reasons, and executes chess moves on a physical chessboard.
• The gantry runs on two NEMA 17 stepper motors controlled by DRV8825 drivers.
• The Raspberry Pi 5 serves as the brain:
  • Uses a computer vision (CV) model with a webcam to detect piece movements.
  • Runs a chess engine to reason and decide the next move.
  • Executes a pathfinding algorithm to determine the best way to move pieces without collisions.
• Instructions are sent to an Arduino, which translates them into motor movements.
• Chess pieces are magnetized and moved using an electromagnet attached to the gantry.
• On the software side, the web app integrates AI agents, coaching features.

⚡ Challenges we ran into

• The two biggest challenges we can into were building the 2nd axis of the gantry, and the Computer Vision system to detecting chess pieces.
• In addition, calibrating the pathfinding algorithm for smooth, collision-free piece movements was tricky and time-consuming.
• Integrating multiple hardware and software components (Pi, Arduino, motors, CV, chess engine) into a seamless system.
• Our two most significant issues took up to majority of the hackathon period to solve, and only worked as we worked simultaneously and collaborated

🏆 Accomplishments that we're proud of

• The finished product looks good and the 2d gantry works flawlessly.
• Successfully merged AI agents with both the physical board and web app, providing live coaching and guidance.
• Built a system that is both functional and user-friendly, despite the short hackathon timeframe.

📚 What we learned

• Hardware hacks require significant planning, testing, and take more time than anticipated.
• Importance of team collaboration and parallel problem-solving to meet deadlines.

🚀 What's next for ChessMate

• Adding a built-in camera system so that remote players can appear via webcam, allowing the physical player to see and talk to them directly during gameplay.
  
• Exploring self-learning AI that adapts to a player’s style and provides tailored coaching.

Built With

• agents
• ai
• arduino
• groq
• nextjs
• python
• raspberry-pi
• react