Skip to content


Interviews, Media Coverage, and Deep Learning

There has been some nice media coverage of our Tensegrity Robotics research “recently”. I’m a bit behind on updating this blog, so “recent” really means “in the last 6 months” or so.

Most exciting is an interview by Zygote Quarterly. This is a magazine dedicated to bio-inspired design, and the interview gives some nice insights into how I’ve found inspiration in fusing engineering and biological inspiration in developing the field of tensegrity robotics. And since it is a design magazine, its really pretty with some nice inspirational photos. Worth the read!

Next, back in January the BBC released a nice short video about our research. I always appreciate when someone does independent research and pulls together a story including prior videos and information we have released. It is so refreshing to see given how much of online media is slapped together with minimal effort and is often wildly inaccurate, especially around science topics. As usual, the BBC continues to maintain standards for real content. Thanks!

And, now to give this post some more fun technical content — back in May we presented a Tensegrity Deep Learning paper at the International Conference on Robotics and Automation (ICRA) that was developed jointing with our colleagues at UC Berkeley who led the algorithm development. In it we demonstrated the first example of learned continuous locomotion on the actual SUPERball robot. What is particularly amazing is that we were able to do all the training in simulation, and have the learned policy work directly on the real robot. This is unusual in the world of robotics, and I believe that it highlight the value of flexible, compliant robots like our tensegrity robot — the compliance that makes it resilient to unexpected contact also makes it robust to approximately tuned controllers. You can read the paper, find the code, and more on the project website.

Posted in Robots, Tensegrity.

Tagged with , , , , , .


My 2016 NASA Ames Summer Series Presentation: SUPERball: A Biologically Inspired Robot for Planetary Exploration

Back in June of 2016 I was very honored to be invited to present as part of the NASA Ames Summer Series. Each summer, the Office of the Chief Scientist at NASA Ames produces a lecture platform with leaders whose high achievements generate innovative discussion, as well as inspire and catalyze scientific progress. This year, the Summer Series consisted of 18 seminars by lecturers from NASA Ames Research Center, external NASA staff, as well as renowned colleagues who lectured on topics that span across multiple advanced subject areas including space technology and space exploration. It was an honor to be included in the lecture series!

Abstract:
Exploration and Innovation both require bold leaps into the unknown, beyond the boundaries of current knowledge and experience. Exploring the unknown frontiers of space requires resilient and adaptable robots capable of surviving the unexpected, qualities which humans excel at. Moving beyond the traditional designs for rigidly constructed fragile robots, Vytas draws inspiration from the flexible tensile network of muscle and tendons of our bodies to develop a new class of “Dynamic Tensegrity Robots.” His current project, SUPERball, is intended to survive high-speed landings without an airbag, and thus enable exploration of treacherous terrains where slipping and falling is an unavoidable possibility. These new robots break the rules of traditional robotics engineering, requiring innovation at all levels of mechanical design, actuation, sensing, and control strategies. Modern neuroscience provides insights into how decentralized rhythmic controllers can enable self-organizing control strategies for this new class of biologically inspired robot and provides insight into our core human qualities of thought, motion, inspiration, and our essential ability to see connections between people and ideas which is at the heart of innovation.

Biography:
Vytas SunSpiral is an entrepreneurial researcher moving fluidly between leading startups and building research labs to explore cutting edge robotic and AI technologies. He is a Fellow of the NASA Innovative Advanced Concepts (NIAC) program,, and currently leads the Dynamic Tensegrity Robotics Lab (DTRL) within the Intelligent Robotics Group at NASA Ames Research Center. His research spans a multi-disciplinary fusion of robotics, physiology, AI, mechatronics, and neuroscience, with the goal of understanding human intelligence via the foundational role that motion plays in our evolution. This quest led to a fundamental new approach to robotics that has the potential to reinvent how we explore the solar system. He is an author of ~50 journal and conference articles and was a contributing author of the 2013 Roadmap for US Robotics. Over the last 20 years he has also been the Founder, CTO, and Advisor to multiple startups, including Mobot, which sold the worlds first commercially available autonomous tour guide robots. Vytas holds a Masters in Computer Science and a BA in Symbolic Systems from Stanford University.

Posted in Bodies, Brains, Presentations, Robots, Tensegrity.

Tagged with , , , , , , , , .


New Video Highlighting Jonathan Bruce’s Research

Jonathan Bruce is a UC Santa Cruz PhD student who has been a central member of my lab for many years now. He has been instrumental to all of our successes, and is a gifted mechatronics engineer who is pushing the boundaries of tensegrity robotics design and control. The UARC program is a mechanism by which UC Santa Cruz and other UC campuses collaborate with NASA Ames, and they recently choose to highlight his innovative work and contribution to NASA’s research success. They sent a film crew to interview him and get context on our broader research project, including our collaboration with the BAIR and BEST labs at UC Berkeley. The resulting video is a *really* excellent overview of our research and well worth watching.

Finally: A Big THANK YOU to Jonathan for all his contributions!
Check out the many Journal and Conference publications he has led or co-authored.

Posted in Robots, Tensegrity.

Tagged with , , , , .


Related Projects by Collaborators

Happy New Year!
I would like to take this opportunity to highlight the tensegrity robotics related work by some of our many inspired collaborators. First of all, I’m excited to introduce Dr. Julian Rimoli of Georgia Tech, who has developed some excellent new tools for analyzing the structural response of a tensegrity robot when it lands on another planet. The resulting video of its dynamics is excellent.

This is what Julian has to say about his work:

“Most approaches to modeling tensegrity structures assume their bars are rigid, and that they only experience pure axial loads. In addition, a common design constraint is assuming that the structure would fail if any of its members buckle. The first two assumptions break down under highly dynamic events such as impacts, and the third one is not necessarily true: slender bars can sustain a load after failure, and consequently stresses would redistribute without necessarily producing structural failure. This video shows an example of a light-weight tensegrity structure under a highly dynamic event. The model accounts for the body forces and associated bending on bars, and their buckling and post-buckling behavior. The ground is modeled as elastic with friction. For those interested, details of the model will be presented at SciTech in January 2016.”

Next, Julian made a great video from the perspective of a camera mounted at a centrally suspended payload during the same landing event as the video above. This shows what the point of view might be for navigation purposes if you gimbaled the camera to stay stable while the robot bounced and rolled. This is a 360 degree video, so you can use the arrows to change the direction that you are looking out from the robot.

Finally, another video of his shows how waves of landing forces might propagate in an interesting manner through the tensegrity robot, making it appear to “inch-worm” its way back up into the air. Once again, this shows how unique and surprising these structures can be!


Next I would like to share the work of a team led by Will Buchanan who built an amazing tensegrity art structure and took it out to the Burning Man festival, where folks could climb and play on it.

Two members of our lab were part of the effort — Ken Caluwaerts and Atil Iscen, both of whom have contributed to the design and control of our SUPERball robot here at NASA.  Best of all, Will and his team captured all the details and lessons learned from creating this giant tensegrity sculpture into an Instructable.  Now you can go make your own giant tensegrity sculpture and have your friends play on it!


Finally, I would like to highlight some recent work by Ryan Adams, who has been an amazing contributor to the development of our open source physics based tensegrity robotics simulator (NTRT — the NASA Tensegrity Robotics Toolkit).  Inspired by our use of coupled oscillators in the controls of tensegrity robots, he has been exploring the dynamics of fields of coupled oscillators.  What is fascinating is how stable dynamic patterns readily emerge out of a randomly seeded field.  This is not just the simple case of all the oscillators synchronizing with each other, but rather the emergence of stable repeating complex patterns, as shown in the video below.

OSim Demo #1 (64×64 grid, 2d) from Ryan Adams on Vimeo.

I think that this is a very important line of research for the understanding of neuroscience and the fundamentals of how we control motion — where a coordinated set of actions must be generated by a very noisy and error prone computational system (our neurons). This ability to start with a random set of oscillators, and have it settle into a stable behavior is exactly what would enable the robust and reliable behavior of animals despite the noisy reality of our neurons. This is obviously just an early stage exploration of the key principles, and a long way from a full theory of neuroscience, but it is valuable to see the key properties at play.

Posted in Robots, Tensegrity.

Tagged with , , , , , , .