MATLAB Robotics System Toolbox and ROS

By Ilia Baranov

Recently, Mathworks released a new toolbox for Matlab. This is exciting for a number of reasons: it includes everything from data analysis to coordination of multiple robots. In today’s post, we explore using this Robotics System Toolbox to connect to both real and virtual Jackal robots.

The toolbox has made a number of improvements since the “beta” version that we wrote a tutorial on a while ago. Matlab now supports services, parameters, analyzing rosbag data, and has a very robust series of tutorials. They even support generating code in Matlab Simulink, and then having it run on a ROS robot, with no extra downloads needed. This should make development of control algorithms faster for robots, and enable fairly detailed testing outside of ROS.

The following video looks at running a Jackal in circles, in both real and virtual space. It also compares a simple obstacle avoidance program, using lidar, that wanders around the environment.

To run the obstacle avoidance sample on a Jackal, ensure that the Matlab Robotics System Toolbox is installed, and download this file: Jackal_New.m

Change the first IP address to your Jackal (real or simulated) and the second to your own computer (the one running Matlab).

If you are ready to try analyzing bag data, download our sample file from the run shown in the video, and plot the laser data.

laserAvoid_Real.bag , scan_plot.m

Link to the toolbox:

http://www.mathworks.com/products/robotics/

Looking for other Clearpath tutorials? Check these out.

Clearpath Robotics Clearpath Robotics is dedicated to automating the world's dullest, dirtiest and deadliest jobs through mobile robotic solutions.

Clearpath Robotics Clearpath Robotics is dedicated to automating the world's dullest, dirtiest and deadliest jobs through mobile robotic solutions.