Controlling Kinova robots with MATLAB toolbox and existing MathWorks code

Controlling Kinova robots with MATLAB toolbox and existing MathWorks code

(Ist derzeit nur auf Englisch verfügbar)

During our stint as FAE (Field Application Engineers) interns with the Kinova Innovation team, we worked on an assignment using the MATLAB programming platform. The goal was to use this new tool in order to control the KINOVA®Gen2 Ultra lightweight robot with the API C++ functions.

Simon Michaud, Samuel Hovington, Students in Robotics Engineering at Université de Sherbrooke

Objective: Augment the robot’s functions by combining them with those from MathWorks

Controlling the KINOVA®Gen2 Ultra lightweight robot with MATLAB allowed us to add MathWorks functions to those supplied by Kinova. We wanted to use this to demonstrate that the robot is enhanced by MATLAB, and that Kinova’s robotics solutions easily adapt to external programming platforms.

We started by exploring how to use the MATLAB precompiled programs to access the C++ functions of the robot API. A procedure was established to open the precompiled folders in various programming environments and then use their contents.

The great news: A user wishing to use MATLAB to control Kinova robots, but having part of their application code in other programming environments, can migrate their existing work to the MATLAB environment.

“Researchers can use MATLAB to control Kinova robots and even migrate their existing work to the MATLAB environment.”

Simon Michaud, Samuel Hovington

Getting to know the MATLAB robotics toolbox

In order to create our presentation, we familiarized ourselves with the MATLAB robotics toolbox. Among the wide range of elements offered by the toolbox, we decided to use the machine vision capabilities. With this, we were able to have the robot move around to find specific objects among several others and have it stack them in a sequence, before eventually repositioning them in their original locations.

This demonstrates that Kinova’s robot adapts to different programming environments, which allows researchers to spend more time on their specific applications rather than on integration of the robot. Thus, users benefit from a robot that is flexible and modular, is easy to integrate and saves them time.

Researchers benefit from a robot that’s flexible and modular, is easy to integrate and saves them time.

More great news

We opted to use the MATLAB robotics toolbox, more specifically the part dealing with computer vision, but the quantity of other resources available is also very impressive, and can save duplication of effort on robotics issues which have already been solved. For example, the toolbox would enable the integration of the KINOVAGen2 Ultra lightweight robot on a mobile platform, since it includes the algorithms required for mobile applications.

All things considered, the use of such tools adds significant value for Kinova’s employees since they allow them to innovate and look for new ways to optimize the robotics solutions already developed.