OnQ Blog

Open Source Robotics Foundation extends support to Qualcomm Snapdragon Processors

3. Dez. 2014

Qualcomm products mentioned within this post are offered by Qualcomm Technologies, Inc. and/or its subsidiaries.

Qualcomm was recently the platinum sponsor for the Robot Operating System (ROS) developer’s conference, ROSCON 2014. At the conference, Open Source Robotics Foundation (OSRF) announced it will extend Robot Operating System (ROS) support to Qualcomm® Snapdragon™ processors. ROS is an open source robotics framework which provides a collection of tools and libraries that simplify the task of creating and programming robotic platforms and applications. It has enabled robotics developers to build upon each other’s work thereby accelerating robotics research and product development. Enabling ROS on Snapdragon processors will allow existing ROS developers to use already existing and future Qualcomm technologies that have been built for mobile. 

With the agreement between OSRF and Qualcomm in place, developers will be able to easily install ROS packages via apt-get for Linux on Snapdragon 600 based board such as SBC IFC6410. For Android, selected ROS packages will be cross compiled which can be used with ROS on Android. This will enable mobile phones to be the main computing device on a robot.

One of the main questions ROS developers have is if Snapdragon processors are powerful enough to run computationally intensive algorithms in ROS. At ROSCON 2014, Qualcomm showed demos of the ROS mapping and navigation package (which are some of the most computationally intensive algorithms) running on Snapdragon 600 based development board (IFC6410). The existing Turtlebot was used for demo where the Intel based laptop was replaced with IFC6410.

This demo proves that Snapdragon 600 processors have adequate computation power to run computation intensive algorithms in ROS.

Qualcomm additionally showed how ROS developers can use existing Qualcomm libraries like the FastCV™ computer vision library, which has optimized performance on Snapdragon processors. 

In this demo, algorithms in FastCV libraries (also available for Linux) is used to track the red ball and ROS packages are used to drive the robot.

These demos give a peek at what Qualcomm has to offer to the ROS community. Qualcomm is excited about its collaboration with OSRF which will act as a catalyst in building robots using superior mobile computing platforms.

Also, I wanted to let you know that we’re looking for startups to join the Qualcomm Robotics Accelerator, powered by Techstars, a 16 week program held in San Diego, which will welcome the inaugural class in May 2015. If your robotics startup could use help with access to hardware, funding and mentorship, apply by February 22, 2015.

Engage with us on