Yertle
A 3D printed quadrupedal robot with 4 limbs of 3 degrees of freedom each, built for locomotion research and reinforcement-learned walking.
- Quadruped
- Open source
- Works with AI
- ROS 2
- 3D printable
Suggested by Robotheca, created by @Jerome-Graves (opens in a new tab), from github.com/Jerome-Graves/yertle (opens in a new tab). Pinned to commit 9d04e2d (opens in a new tab).
- Hardware license
- MIT (opens in a new tab)
- Software license
- MIT (opens in a new tab)
- Commercial use
- Allowed
This repository contains executable files that Robotheca doesn't check. Only run them if you trust the creator.
Which files (1 script)
Photo and 3D model
3D preview converted from github.com/Jerome-Graves/yertle at 9d04e2d, under MIT. View the original files from GitHub (3.4 MB) (opens in a new tab)
- Reach
- 20 cm
- Weight
- 1.8 kg
What it can do
Written by Robotheca from the project's documents.
- Walks as a quadruped, with four legs of three servo-driven degrees of freedom each.
- Hand-tuned sinusoidal gait controllable from a Python GUI with arrow keys.
- Learned walking policy trained with reinforcement learning, tracking commanded body velocity.
- Trained policy can run as a ROS 2 node that publishes twelve joint targets.
- Simulation available in PyBullet and NVIDIA Isaac Lab, including rough-terrain training and distillation to a sensor-only policy.
Works with AI
Found by Robotheca in the project's files at commit 9d04e2d, its robot.yaml and Robotheca's notes on its documents. Never guessed; Robotheca doesn't run the code.
Uses reinforcement learning: rsl_rl
In the project files: isaac_lab/rsl_rl_distill_cfg.pyUses reinforcement learning: Isaac Lab
In the project files: isaac_lab“Learned walking policy trained with reinforcement learning, tracking commanded body velocity.”
Robotheca's notes: What it can do
Bill of materials
This repository has no parts list Robotheca can read yet.
Build guide
The guide appears here as you scroll.
Guide by @Jerome-Graves (README.md), under MIT.
More guides
Wiring and electronics
Notes written by Robotheca from the project's documents.
Main parts
- ESP32 microcontroller handles sensors and servo control, and communicates over WiFi or serial.
- Raspberry Pi 4 single-board computer is optional and runs vision, localisation, ROS and advanced controllers.
- Hobby servos, three per leg, driven through a PCA9685 servo driver.
- MPU9250 9-axis accelerometer and gyro sensor, optional.
- Optional current and voltage sensor.
- 7.4V 2S LiPo battery of about 5000mAh, with a 6V SBEC regulator.
Connections
- Servo driver is connected to the ESP32 to drive all twelve servos.
- The ESP32 is mounted on top of the Raspberry Pi, with a custom Pi hat that also carries the IMU.
- The battery feeds the servos through a 6V SBEC power regulator.
- Original Kangal and SpotMicro wiring can be reused with this robot.