🤖 Robotics-Controlled Chess and Drawing System with QARM Robotic ArmR

Overview

This project introduces an advanced chess-playing robotic arm system, which leverages distributed architecture, intelligent decision-making, and precise control techniques.(Patent Pending, Application No. 202510598609.1, 1st Inventor).

Below are three videos showcasing the robot’s capabilities: human-robot chess interaction, drawing a Christmas tree, and creating a heart.

Project Highlights

Designed and implemented a QARM four-axis robotic arm system to enable human-robot chess interaction and drawing tasks.
Developed control and decision systems integrating multiple platforms:
- MATLAB for workspace simulation and initialization.
- Simulink for robot control logic, inverse kinematics, and sequence logic design.
- Visual Studio for decision-making algorithms utilizing a pre-trained chess AI.
Innovated a magnet-assisted gripper and a passive leveling tool using gravity to ensure horizontal placement in under-actuated operations, enhancing accuracy and compatibility with robotic arms of varying DoF.
Applied plane fitting algorithms to correct robotic arm positional errors, ensuring accurate precision.

Flowchart: Decision and Motion Process

Below is a simplified flowchart describing how various modules interact:

graph TD A[User Interaction] --> B{AI Decision System} B --> C[Simulink Control] C --> D[Actuate Arm] D -->|Check Result| E{Success or Retry?} E -->|Success| F[End] E -->|Retry| B

Challenges and Innovations

Key Challenges:

Precision Issues:
Initial central alignment strategies led to XY and Z errors due to under-actuation.
Solution: Plane fitting using calibration points and fitted matrix algorithms.
Magnet Control:
Balanced magnetic forces between chess pieces and the board, ensuring stable pickup and release.
Integration Limitations:
Bridged communication challenges between MATLAB, Visual Studio, and Simulink using TCP/IP communication，CSV files and memory-based data sharing.

Innovations:

Passive Adaptive Gripper: Maintains a horizontal placement for under-actuated end-effectors via gravity-driven rotation.
Distributed System Architecture: Partitioned decision-making, control, and vision across three independent modules for enhanced scalability.

Thank you for exploring this project. Feel free to share your thoughts! 🙌