Mark Micire (research scientist at the Intelligent Robotics Group at NASA Ames) and Yun Kyung Kim (human-robot iInteraction designer at NASA Ames) were incredibly generous in offering me an opportunity to speak with the AstroBee and Super Ball Bot groups at NASA Ames. We’ve been keeping an eye on Super Ball Bot over at Super-Releaser, particularly because of the way the teams working on it are bringing simulation and iterative prototyping together to solve the open-ended problems involved in designing a robust control system for bots that can configure themselves into nearly infinite shapes.
The talk focused on the opportunities to use compliant materials to replicate organic mechanisms, the ways Super-Releaser solves problems in soft robotics, and the way we integrate multiple disciplines into our research. Afterwards I was able to see the work of the Super Ball Bot team – developing novel compliant actuators in addition to refining the systems that power their current Ball Bot prototypes.
I was also able to see the AstroBee, which was being evaluated on the biggest granite surface plate I’ve ever seen. I got to talk with Yun about her experience as a designer integrating into a team of engineers, which is its own challenge in itself, and the goals of the AstroBee project. It’s going to serve as a platform to develop behaviors for human/machine interaction in 0g, which is a problem I’ve never even considered.
Kari Love and I gave a talk at Maker Faire last year detailing how the maker mindset (tinkering to get an intuitive sense of the rules governing the system, hands-on learning, fast frugal iteration, and sharing) can be transformative for research into fundamental technologies and chronically intractable problems.
The key factor is going from zero to a working understanding of the ground truths underlying the problem you’re trying to solve as quickly as possible. From historical surveys of how transformative technologies have been developed in the past (like TRIZ), deeply focused research is no match for playful learning and interdisciplinary exploration.
These are the techniques we use at Super-Releaser to get things done given how new the field is and how much it relies on an intuitive understanding of the mechanics of soft systems. When there isn’t a robust framework to simulate before experimentation, you need to rely on experience and spot tests.
We were also very proud to have our intern, Aidan Leitch, give his own talk on his soft robotics research. It was very well attended and people seemed excited to see live demos of his soft robot designs.
Super-Releaser has begun work on a book on soft robotics for Maker Media. Kari Love and I are writing a book that provides a history of the field of soft robotics, tutorials demonstrating its basic principles, more sophisticated projects like a control system and entire soft robots, and the potential of applied soft robotics from medical devices to human spaceflight to interplanetary exploration. As far as we can tell this will be the first book published demonstrating practical soft robotics.
We are working with Roger Stewart to complete the text before the end of this year. Fingers crossed it will be available in bookstores in early 2018.
Jacob Alldredge invited me to speak at APL to speak with their research staff as part of their REDD Talks series. I presented a talk on the research process Kari Love and I developed at Super-Releaser for rapidly evaluating and developing novel technologies: The Physical Feedback Loop.
It was encouraging speaking with scientists and engineers working at the leading edge of their fields about how they picture their own research processes, and how they tackle problems in novel areas. I got some fantastic feedback from project leads at APL, and was sincerely impressed by their internal manufacturing processes which produce everything from novel 3d printed metal compounds to NASA satellites.
I’ve been going to CCC for years, but this is the first time I’ve gotten a talk accepted in one of the main venues. It was thrilling to share my research with such a wide audience. I spoke about the kinematics of soft bodied organisms, designing soft robots, and future applications for compliant mechanisms. Below is a complete video of the talk and the Q&A session afterwards.
Yesterday I gave a talk about incorporating soft robotics, compliant mechanisms, and biomimetic structures into your engineering toolbox at NYU. I’ve been interested in how compliant mechanisms can reduce the computational complexity of tasks like manipulation and locomotion and this talk was a good opportunity to share some of my ideas on the subject.
The general thesis is that biology presents a huge trove of solutions to problems in robotics especially directed at optimizing the amount of sensing you’re devoting to understanding an environment and the amount of computation you’re devoting to navigating that environment. Compliance is an essential tool for creating systems that reduce a wide range of potential inputs into a simplified space of positive outputs.
Case in point:
You can find my slides here. If an audio/video copy becomes available I will update this post with a link.
When I spoke at the SpaceApps conference, I hadn’t realized how close I was to working with NASA in a much more official capacity. A few months earlier I developed some prototypes for Final Frontier Design, a company devoted to the design and engineering of spacesuits. This was in my role as lead scientist at Super-Releaser and the end goal was proving to NASA that mechanical counterpressure garments (like I described in my talk) could be a practical reality with some time and development. I’m pleased to announce that they approved our proposal and we will be working on a new generation of EVA gloves over the next six months.
What I’m most excited about is the opportunity to bring all of the elements of engineering, prototyping, and digital manufacture for compliant materials to create and test all of the iterations of the glove we’ll be going through. There are thrilling mechanisms and intricate problems that I believe are only workable with compliance in mind. After all, we’re interfacing with the most mechanically complex manipulator the body has to offer.
Hopefully I’ll be able to update as progress is made. I will still continue my development work at Super-Releaser and research on the Neucuff, though the frequency of updates may drop off.
The Glaucus is a soft robotic quadruped composed of a single seamless silicone part. It has a complex network of interior channels, created via a lost wax process, that turn into actuators when pressurized with air. It’s able to walk with a diagonal gait, similar to a gecko or Glaucus Atlanticus sea slug, using only two input channels.
The Glaucus was created to demonstrate a method for fabricating soft robots of nearly any geometry with arbitrary interior structures. It’s been my goal, since beginning my research into soft robotics, to simplify the process of prototyping and refining designs. Often the barrier between an interesting bench prototype and practical application is how it scales into production. If methods for experimenting with the core concepts, evaluating them in a context that represents their final manufactured state, and refining them for mass production don’t exist, the idea is very likely to languish on the bench. Continue reading →
I was hired by SOLS to help out with their Adaptiv project. The idea was to showcase the procedural modeling techniques, materials, and technologies behind their printed insoles with a futuristic robotic shoe. Jordan Dialto, the industrial design lead at SOLS, approached me in my capacity as lead scientist at Super-Releaser to make a prototype soft robot shoe that could change shape and fit in response to the wearer.
The project started out with an external shell modeled by Continuum Fashion. Although the design was elegant, this posed a challenge for introducing the robotic elements and the engineered components that would stitch everything together. Since the external shell was generated in a mesh CAD program, it didn’t fit into SolidWorks’ reference frame. This meant using the mesh as a reference and generating a simplified surface to extrude the soft robot elements and retaining skeleton from. Continue reading →