Joanne Pransky, associate editor of Industrial Robot, recently interviewed Dr. Robert Ambrose, chief of the Software, Robotics, and Simulation Division at Johnson Space Center in Houston. Ambrose was interested in working for the National Aeronautics and Space Administration ever since watching the Apollo 11 mission as a child.
Now, he supervises on-orbit robotic systems for the International Space Station, the team for tele-robotics and autonomous systems, and the human spaceflight architecture study team at NASA. Among other things, Pransky and Ambrose discussed his educational experience, past and current robotics projects at NASA, and the prospects for human spaceflight and engineering students.
This interview is available free to Robotics Business Review readers until Sept. 30, 2015. Here’s a preview:
Pransky: What has been your favorite project that you’ve worked on thus far in your illustrious career and why?
Ambrose: I’m convinced my favorite hasn’t happened yet, but it’s out there. I’ve been very fortunate to work on some really good projects. The first project I had when I came to JSC was working on some robot joints inside a thermal vacuum chamber. I learned about the challenging environment in space, and that has really had an impact on me throughout my career. The harshness of vacuum and working in extremely hot or cold environments has permeated all my thinking.
For my next project, I was sent around the country to do a technology capture on a very large sophisticated robot in which the build had just ended. My job was to ensure we didn’t lose any of the lessons learned. I met with every vendor that had built a component for the robot: the motor vendors, the bearings, the sensors, the cable harness vendors, etc. The robot, which I have since kept in a vault, never flew, but it taught me a lot. Out of the ashes of that project were all the tough lessons that stood on the team’s shoulders when we designed Robonaut.
Pransky: When did the Robo-Glove project start and how?
Ambrose: When we started the partnership with [General Motors], we would never have thought of a glove. We started looking at the jobs that would be good for a robot that could work safely next to people and prevent repetitive-motion injuries. NASA’s interest is just to get the technology out there so that it helps society here on Earth. For example, we were contacted by the construction industry to pick up dangerous materials such as a 4-by-8 sheet of metal.
We thought that if we could take the Robonaut hand and wear it as a glove so that the glove is squeezing the object for humans, then the human’s tendons aren’t negatively affected. Imagine a glove that absolutely every time guarantees you will never drop the object. We built Robo-Glove in just two months. We then went through two or three generations of it, and it just kept getting better and better.
A mix of NASA and GM engineers kept tweaking it and playing with different tendon routings and using different sensors so we could program it more like a natural reflex that squeezes upon the touch of an object through a tiny button embedded in the fabric of the glove, and releases when you pull your finger off. Though there are haptic feedback devices out there, there are none that have yet to really conquer applying finger-level forces back to the human hand in a natural way.
We’re now starting a project to integrate that glove into the spacesuit glove. The goal is to make the space suit not as hard on the astronauts’ hands. We’re looking at adding in some actuators that would wrap around the fingers of a glove that are pressurized. Hopefully, that will make it easier for the astronauts.
Pransky: What is NASA’s vision for the complete Robonaut 30 years from now?
Ambrose: My personal vision is that Robonaut is a 50-year project. We started the Robonaut 1 Project back in 1996, 19 years ago.
I was the only one who called it the Robonaut 1 Project, by the way. The first one we built, I called it Robonaut 1A. (R1A is now in the Smithsonian, very patiently waiting for future Robonauts to come visit.) We built two units from 1996 to about 2006.
The second one was portable. We put it on Segways. We put it on four wheels. We had a single leg, and it did climbing with its arms. That’s where I learned that we didn’t learn to climb with our hands. We learned a lot from Robonaut 1, and by 2006, we had all these pent-up ideas. That’s about the time I gave the seminar to GM, and they said, “We like some of those ideas. Let’s get together.”
That’s when we started the Robonaut 2 (R2), which was pretty much a clean sheet. GM then moved their engineers to Houston for 24 months for the design and build. It was a great, aggressive team working together in one lab. At 12 months, we had built the R2A right and left limbs. At 18 months, we had the full two arms with a torso and a waist. At 24 months, we had the second unit, R2B, which went to GM.
Then, we had an opportunity to go to the space station. The head of the space station asked in January, “Can that robot go? Can it be at the Cape packed in a shuttle by July?” We said, “Yes.” But we had to take the better one, which was the one owned by GM, so we took theirs back, and R2B was the one that we sent to the space station. We ended up building GM a replacement, R2C, which was even better….
More Essential Interviews:
- Helen Greiner, CEO of CyPhy Works
- Esben Ostergaard, CTO of Universal Robots
- Rodney Brooks, Founder of Rethink Robotics
- Nest’s Yoky Matsuoka
- iRobot’s Paolo Pirjanian
- When Bill Townsend Met Burton Doo
- Mark Tilden, Inventor of BEAM Robotics
- Rich Walker, Shadow Robot Company
- Russ Angold, Co-Founder, Ekso Bionics