Robotic Chess Project: AI Vision Model & Code (UR3e) COMPLETED

*** IMPORTANT: COMPLETED ***

This is a live project and a prototype version.

Build your own robotic chess opponent! This is an university project to create a complete system that uses a UR3e robot arm to move pieces on a real-world board.

ALL PYTHON SCRIPTS ARE AVAILABLE: NOW!

The heart of this project is a robust, pre-trained computer vision (CV) model designed to identify all chess pieces. It has been optimized to operate in Black & White (B/W), which dramatically improves detection reliability and speed—a key requirement for real-time robotics.

With the provided Python scripts (when released) and your hardware, you can bring this incredible automated game system to life.

Technology Stack:
Vision (Included): Pre-trained CV model.
Control (Included): Python scripts to integrate vision with the robot.
Required Hardware: UR3e Robot Arm and a 3D Printed (STL) Chess Set.
Camera: A webcam is required (the Logitech C920 is a good option).
Board: FDM (28x32 cm)

Project Roadmap
Follow the progress of this exciting project!

✅ COMPUTER VISION [COMPLETE]
The model correctly identifies pieces and their positions.
✅ PIECE & BOARD DESIGN [COMPLETE]
Designing custom 3D printable pieces and a board with a defined robot-safe zone.
✅ ROBOTIC CONTROL [COMPLETE]
Developing kinematics and gripping routines for the UR3e using Python.
✅ CHESS AI [COMPLETE]
Integrating a chess engine (AI) for decision-making.

Impresora FDM
0.28 mm
130 mm/s
240°C
PETG Obligatorio