Robotics & Vision Toolboxes for Python

IEEE RAS School Reproducible Research Performance Evaluation and Benchmarking in Robotics

Lecturers

Program

Lecture (10am CET on 27 January 2022)

This session will give you an introduction to some Python toolboxes that make it easy to model and simulate quite complex robotic and robotic vision systems. We will cover the following topics:

transforms, quaternions and all that stuff (Peter), 8 mins
robot manipulator arms; cameras (Peter), 8 mins
cameras (Peter), 4 mins
Swift - a fast browser-based simulator (Jesse), 20 mins

A paper at the most recent ICRA gives a gentle introduction to the Toolboxes and our design philosophy.

PDF slides available here:

Peter's slides (PDF)

Hands-on workshop (11am CET on 27 January 2022)

We provide Jupyter notebooks that you can run with zero-install on your computer (apart from the Swift simulator which can't work with CoLab). These reinforce the material covered in the lecture.

icra2021.ipynb this takes you through core features of spatial mathematics and arm robots
swift.ipynb this takes you through core features of the Swift simulator and arm-based robotics
visualservo.ipynb is an example of image-based visual servoing of a manipulator arm. It is for you to complete, but we have sketched out some of the solution.

How to get started with the code

Local install

This is the recommended way to experience this workshop. With the local install, you will be able to experience the Swift Simulator and related excercises.

Install conda following instructions here https://docs.conda.io/en/latest/miniconda.html

// Make a new conda env
conda create -n 'ws-rtb' python=3.9

// Activate the new conda env
conda activate ws-rtb

// Install the toolbox and deps
conda install -c conda-forge roboticstoolbox-python==0.11.0
conda install -c conda-forge swift-sim==0.10.0

// Install jupyter notebook 
conda install -c conda-forge notebook

// Clone this repo to operate locally
git clone https://github.com/jhavl/ws-rtb.git

// Change into the directory
cd ws-rtb

// Open Jupyter in the current directory
jupyter-notebook

// pip install alternative
pip install roboticstoolbox-python==0.11.0
pip install swift-sim==0.10.0
pip install sympy
pip install notebook

Using CoLab

Colab provides an install free way of participating in the workshop, although not all excercises will be compatible.

Go to CoLab and choose the notebook you want to run. The two that work (mostly) with CoLab are

icra2021.ipynb this takes you through core features of spatial mathematics and arm robots
visualservo.ipynb a classic image-based visual servo system (IBVS) implemented with the Toolboxes, and an assignment for students