Abstract:

Mobile robotics has surged in popularity with the emergence of applications for commercial and industrial use such as autonomous cars, drones, or warehouse robots. Accordingly, new practitioners versed both in the theoretical and the practical aspects of the field are on high demand. Supporting material is key for the training of such stakeholders, especially in a multidisciplinary field such as robotics, which includes numerous and heterogeneous concepts from mathematics, statistics, physics, etc. In this regard, these complex concepts are better understood when presented in the context in which they are applied, where the environment, goals, and actions can be clearly visualized. This fact calls for modern and quality material using methodologies and tools applied to real use cases, seamlessly introducing theoretical concepts and their implementation and moving away from the traditional theory/exercise pedagogical approach.

This paper presents a collection of educational Jupyter Notebooks for use in undergraduate robotics courses, which have been built from the ground up to meet these issues. First, they make use of the Python programming language, praised both for its ease of use and the breadth of its library support. It has gained particular relevancy in Computer and Data Science applications, which can be of use in a field increasingly reliant on machine learning approaches as mobile robotics. Second, they are implemented using the Jupyter Notebook technology [1], widely resorted to in well-known online learning platforms such as Coursera [2] or Udacity [3]. Jupyter Notebooks permit us to combine at the same place theoretical explanations through text, images, mathematical equations, videos and/or links to additional resources, as well as executable code, this way producing comprehensive and contextualized material that incorporates the interactivity, dynamic visualizations and possibilities of an application. The notebooks provide a broad view of the robotics field, with a particular emphasis on mobile robotics, as they cover aspects like: probability bases, robot motion, sensing, localization, mapping, and motion planning.

The developed notebooks are meant to be an engaging tool for mobile robotics lecturers, students and practitioners seeking to enhance their knowledge basis. Currently they are being used in undergraduate courses at the University of Málaga (Spain) with promising results. The student version (without solutions) of the presented notebooks is publicly available at https://github.com/jotaraul/jupyter-notebooks-for-robotics-courses, while the complete version can be requested individually by any interested lecturer. This learning tool welcome any contribution from the mobile robotics community.

References:
[1] Kluyver, Thomas, Benjamin Ragan-Kelley, Fernando Pérez, Brian E. Granger, Matthias Bussonnier, Jonathan Frederic, Kyle Kelley et al. "Jupyter Notebooks-a publishing format for reproducible computational workflows." In ELPUB, pp. 87-90. 2016.
[2] Coursera webpage: https://www.coursera.org/
[3] Udacity webpage: https://www.udacity.com/

Keywords:

Jupyter Notebook, Mobile Robotics, Undergraduate Courses, Python.