Loading 3D model… large boards can take a moment.
A comprehensive control system for a six-legged robot (hexapod) using BeagleBone AI (or Black), featuring both kernel-space drivers and user-space applications with reinforcement learning capabilities.
This project implements a complete software stack for controlling a 6-legged robot with 18 servo motors (3 per leg), including:
Kernel-Space Components:
User-Space Components:
Reinforcement Learning:
The system follows a layered architecture with hardware abstraction. For visual reference, see the architecture diagrams in the /docs/diagrams/image directory:
.
├── driver/ # Linux kernel device drivers
├── app/ # User applications and libraries
├── docs/ # Documentation
└── scripts/ # Build and utility scripts
Build the kernel module:
./scripts/build.sh -b driver
Install the kernel module on BeagleBone AI:
./install.sh
Build user-space applications:
./scripts/build.sh -b app
Run the test utility on BeagleBone AI:
./test_servo
See individual README files in subdirectories for detailed development guides:
Individual tests:
./test_servo # Test servo control
./test_mpu6050 # Test IMU sensor
./test_movement # Test movements like tripod, ripple, ...
./test_calibration # Test servos offset for each leg
./test_hcsr04 # Test HC-SR04 sensor
No comments yet. Be the first to ask about this board.