scikit-robot 0.3.14

A Flexible Framework for Robot Control in Python

scikit-robot: A Flexible Framework for Robot visualization and programming in Python

Documentation | Installation | Quick Start | Python API | Contribute

Scikit-Robot is a lightweight pure-Python library for robotic kinematics, motion planning, visualization and control.

https://github.com/user-attachments/assets/1dad1a4d-ba59-4ec0-93b0-96c8718c9e8f

Installation

Quick Install with uv (Recommended)

uv is a fast Python package manager. Create a virtual environment and install scikit-robot:

uv venv
source .venv/bin/activate  # Linux/macOS
# .venv\Scripts\activate   # Windows
uv pip install scikit-robot

Or with all optional dependencies:

uv pip install "scikit-robot[all]"

Install with pip

You may need to install some dependencies by apt-get install:

sudo apt-get update
sudo apt-get install libspatialindex-dev freeglut3-dev libsuitesparse-dev libblas-dev liblapack-dev

Then,

pip install scikit-robot

If you would like to use Pybullet Interface, open3d and fast-simplification for mesh simplification,

pip install scikit-robot[all]

Command Line Tools

Scikit-robot provides a unified command-line interface through the skr command, which consolidates all robot-related tools into a single entry point.

Using the skr Command

You can use the skr command with various subcommands:

# View all available commands
skr --help

# Visualize URDF models
skr visualize-urdf ~/.skrobot/pr2_description/pr2.urdf --viewer trimesh

# Convert URDF mesh files
skr convert-urdf-mesh robot.urdf --output converted_robot.urdf

# Change URDF root link
skr change-urdf-root robot.urdf new_root_link output.urdf

# Calculate URDF hash
skr urdf-hash robot.urdf

# Modularize URDF files
skr modularize-urdf robot.urdf --output modular_robot.urdf

# Visualize mesh files
skr visualize-mesh mesh_file.stl

# Convert wheel collision models
skr convert-wheel-collision robot.urdf --output converted.urdf

# Generate robot class from URDF geometry
skr generate-robot-class robot.urdf --output MyRobot.py

Legacy Commands (still supported)

For backward compatibility, the original commands are still available:

# These commands work the same as their skr equivalents
visualize-urdf ~/.skrobot/pr2_description/pr2.urdf --viewer trimesh
convert-urdf-mesh robot.urdf --output converted_robot.urdf

Visualize URDF

To visualize a URDF model and inspect your robot's configuration:

skr visualize-urdf ~/.skrobot/pr2_description/pr2.urdf --viewer trimesh

If you experience performance issues with the default viewer, try pyrender for smoother visualization:

skr visualize-urdf ~/.skrobot/pr2_description/pr2.urdf --viewer pyrender

For a web-based viewer with interactive joint angle sliders, use viser:

skr visualize-urdf ~/.skrobot/pr2_description/pr2.urdf --viewer viser

Running these commands should open a viewer displaying your robot's 3D model. Below is a sample image of what you should expect.

Viser Viewer - Interactive Motion Planning

The viser viewer opens in your browser and provides a full-featured interface for robot manipulation:

• Joint Control: GUI sliders for manipulating joint angles in real-time
• Interactive IK: Drag transform controls at each end-effector to solve inverse kinematics
• Motion Planning: Plan collision-free trajectories between waypoints with obstacle avoidance

Interactive Motion Planning Demo

The following video demonstrates the motion planning capabilities with collision avoidance:

https://github.com/user-attachments/assets/1dad1a4d-ba59-4ec0-93b0-96c8718c9e8f

Use the viser viewer with motion planning enabled:

from skrobot.viewers import ViserViewer
from skrobot.models import Panda

# Create viewer with IK and motion planning enabled
viewer = ViserViewer(enable_ik=True, enable_motion_planning=True)
robot = Panda()
viewer.add(robot)
viewer.show()

Or via command line:

skr visualize-urdf robot.urdf --viewer viser

Key features:

• Save waypoints by posing the robot via IK
• Add obstacles (sphere, box, cylinder) interactively
• Plan collision-free trajectories using augmented Lagrangian optimization
• Visualize planned trajectory as ghost robots
• Animate the planned motion

Interactive IK Demo

Drag the transform controls at each end-effector to solve IK in real-time:

https://github.com/user-attachments/assets/1294307e-a901-4f5f-aa4e-433cfbfacd03

Create Your Own Robot Model (URDF)

If you want to build your own robot from scratch, you can refer to the How to Create URDF from CAD Software documentation.

Try it in Google Colab!

You can try scikit-robot directly in your browser without any installation:

This interactive notebook demonstrates:

• Multiple robot models (Kuka, Fetch, Nextage, PR2, Panda) with synchronized animation
• Grasp and pull manipulation task with inverse kinematics
• 3D visualization in Jupyter/Colab with smooth updates

Jupyter Notebook Viewer

Scikit-robot includes JupyterNotebookViewer for interactive 3D visualization in Jupyter notebooks and Google Colab:

import skrobot

# Create robot and viewer
robot = skrobot.models.PR2()
viewer = skrobot.viewers.JupyterNotebookViewer(height=600)
viewer.add(robot)

# Display and animate
viewer.show()

for _ in range(10):
    robot.rarm.angle_vector([0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7])
    viewer.redraw()  # Smooth updates without flickering

See examples/notebooks/ for more examples.

Features

• Loading robot model from URDF (examples/robot_models.py)
• Forward and inverse kinematics (examples/trimesh_scene_viewer.py examples/pr2_inverse_kinematics.py)
• Collision detection
• Collision-free motion planning with trajectory optimization (examples/collision_free_trajectory.py)
• Interactive IK and motion planning in the browser (ViserViewer)
• Interactive viewer (examples/trimesh_scene_viewer.py)
• Jupyter notebook viewer (examples/notebooks/)
• Pybullet and ROS command interface (examples/pybullet_robot_interface.py)
• Forward and inverse dynamics