Easier, safer robotic programming is among results from an expanding open-source robotics software group

Cover Story: ROS-Industrial, an open-source factory automation software project, will celebrate its fifth anniversary in January 2017 by expanding into Asia, increasing visibility in a major robotic competition, and is looking to improve workplace efficiency and safety with its programming.

By Chris Vavra December 1, 2016

Open-source robotic programming with increased efficiency and safety are among results noted as the Robot Operating System Industrial (ROS-I) open source software project is about to celebrate its fifth year, noting continued expansion. ROS-I is an open-source industrial robotic software framework and working group established by Southwest Research Institute (SwRI). ROS-I builds on ROS, which is maintained by the Open Source Robotics Foundation. The ROS-I project will celebrate its fifth anniversary in January 2017 by adding a third branch in Asia to complement its U.S. and European organizations, and ROS-I software was used by the victors in a major robotics competition in May 2016. The group has come a long way in five years and show no signs of slowing down.

"We are delighted that things have turned out as well as they have. ROS has brought consensus to R&D robotics across the board. I think we’re only beginning to realize the benefits of that for manufacturing and there’s a lot of hope for the future," said Paul Hvass, the ROS-Industrial Consortium (RIC) Americas program manager at SwRI.

Robotic programming has brought many companies of different sizes and goals together to create a singular language and system that benefits everyone. "Engineering in an open-source way is something you think would never work in theory because of competition and commercial obligations. But it’s great how we’ve been able to harmonize," Hvass said. "A rising tide lifts all boats, and it helps improve the value of automation for everyone." 

Pick-and-place competition

In June, Team Delft competed in the Amazon Picking Challenge in Leipzig, Germany, and won with a robot that was controlled by ROS-Industrial software in conjunction with their deep learning perception system (see Figure 1). The challenge was a global competition with teams from the U.S., Germany, and Japan. Teams were challenged to program their robots to pick objects from totes (stow task) and put them on shelves and also pick objects from stocked shelves and place them in containers (pick task). The robots had to work with a variety of objects and an unstructured environment, which made the tasks more difficult.

"Winning was amazing on so many levels," Hvass said. "Team Delft are avid contributors to ROS, and it was great to personally see them be rewarded for their hard work and ingenuity. The community was very honored with the win." 

Group expands into Asia

The big goal for ROS-I in 2016 was establishing a base in Asia, which has seen a surge in robotics over the last several years. The group’s Asia-Pacific branch formally launched this year and will be managed by Advanced Remanufacturing Technology Centre (ARTC) in Singapore.

Singapore might not seem like the first choice for the Asia-Pacific region, but they are recognized internationally for innovation. "It’s a smaller market compared to China, but it’s globally connected to so many countries, and it’s kind of at the center of where everything is happening with ROS in Asia-Pacific."

Singapore, which Hvass called the Switzerland of Southeast Asia, is also a major investor in the robotic industry. "There’s more public investment in robotics per capita than anywhere else in the world," he said.

The response in the Asia-Pacific region, Hvass said, has been very positive thus far and the group is looking to strengthen its presence. ARTC and Nanyang Technical University (NTU) hosted a sold-out industrial workshop in July 2016 and are looking forward to the 2017 IEEE International Conference on Robotics and Automation (ICRA 2017), which will be in Singapore. 

ROS-I technology developments

Fraunhofer IPA, which leads the ROS-Industrial Consortium in Europe, led the EU-funded ReApp project, which developed an ontology-driven programming IDE that is designed to provide signal flow graph and state flow graph programming tools. It also supports drag-and-drop library nodes, which helps make robot apps easy to find and use. ReApp also helps make robot manufacturing for small-and-medium manufacturers intuitive by overlaying model-based programming on top of ROS (see Figure 3).

The SwRI team developed a QT Creator ROS Plugin, which provides a ROS-centric development environment with ROS component templates. SwRI also is developing a graphical 3-D setup tool called CAD-to-ROS that is designed to make kinematic chains easy to create and modify.

The Factory-in-a-Day project led by TU Delft Robotics Institute also is developing hardware and software components for cost-effective robotization using ROS-Industrial. One development they’re working on is an artificial skin. Each "cell" in this robot skin includes distance, force, and temperature sensors as well as a 3-axis accelerometer that can be used to enable safe interaction with humans (see Figure 4).

"It’s locally smart and it knows where everything is," Hvass said. "It uses signals to change course and we believe it will be a powerful tool from a control and safety standpoint."

Chris Vavra, production editor, Control Engineering, CFE Media, cvavra@cfemedia.com.


Key Concepts

  • ROS-Industrial, an open-source industrial robotic software framework and working group, celebrated its fifth anniversary.
  • The group expanded into the Asia-Pacific region and has established operations in Singapore.
  • ROS-Industrial is developing technology to improve efficiency and allow robots to work safely for humans.

Consider This

What development do you think ROS-Industrial should work on next?

ONLINE extra

See related stories on ROS-Industrial and their past developments linked below.

Author Bio: Chris Vavra is web content manager for CFE Media and Technology.