The Soft Touch of Robotics
Biomedical engineering professor Frank Hammond's research focused on creating wearable, useable devices for humans
When he was a little kid, Frank L. Hammond III would watch the Transformers animated series or, if that wasn’t on, he’d watch Challenge of the GoBots. While both programs were created to help toy companies sell toys (Transformers and GoBots), they managed to do something else for Hammond. They nurtured his budding interest in robots.
“Ever since I was a kid I’ve been into robotics, and these shows featured robots that could change their form, which I thought was very cool,” says Hammond, assistant professor in the Wallace H. Coulter Department of Biomedical Engineering and the George W. Woodruff School of Mechanical Engineering.
“So I’ve always been interested in devices like these, that could change their form or function to fit the needs of the environment,” adds Hammond, who recently joined the multidisciplinary team of researchers at the Petit Institute for Bioengineering and Bioscience.
The overarching goal of his research is right there in the name of the lab he helms: the Adaptive Robotic Manipulation Lab.
“The inspiration for most of my research is getting robots involved in activities of daily life, creating a collaborative environment for humans and robots,” says Hammond. “You’ve seen a lot of the robots that are available commercially – robots that sweep floors for example, robots that are very good at doing a certain job very precisely, quickly and reliably, over and over again for tens of thousands of cycles.”
Instead, Hammond is interested in building machines capable of performing complex tasks in concert with their human hosts, wearable devices that are “adaptive and flexible, in direct contact with humans, capable of augmenting human motion.”
So Hammond shows off a hand-like device with two fingers and a thumb, soft to the touch, like skin – silicone digits formed in a 3D printed mold, actuated through pneumatic means via little CO2 cartridges.
“This is biologically inspired,” Hammond says. “And it’s a little safer than the rigid, electrically-powered devices. This mechanical hand won’t smash the desk if you move too quickly. It’s powered by air and is mechanically compliant, so it’s not going to damage the wearer or the environment. It’s too soft to do that.”
The replacement hand can grasp a coffee cup or a bottle of water, open a door, and what it may lack in precision (at this point), it makes up for in utility and cost. One of these devices costs less than $100 to make.
“I think the economy of this device, in general, is closely intertwined with the proliferation of self-robotic devices,” Hammond says. “Soft robotics is much more economical now that so many researchers and companies are doing it. Prices have fallen and there’s more competition.”
Hammond demonstrates the hand and its pneumatic actuator. It’s a self-attaching device – you can attach it using one able arm. He lowers air pressure or changes the air distribution to control how the fingers bend and move. And though Hammond isn’t saying his robotic hand is a full-on replacement for your real hand, it does offer the user some options that Mother Nature can’t.
“These are modular components,” he says. “So for example, if I wanted to have longer fingers, I could very easily swap these out with longer fingers.”
Hammond earned his Ph.D. at Carnegie Mellon, where he worked and studied in the department of mechanical engineering and The Robotics Institute. He did his postdoctoral work at Harvard and the Massachusetts Institute of Technology (MIT), where leading researchers are directing a number of projects aimed at giving humans extra capability.
Robotics researchers have coined a phrase, “supernumerary robotics.” Imagine having extra limbs (supernumerary robotic limbs, or SLRs), which are not designed to replace a missing limb, but to augment what you have.
“They’re still using rigid linkages and electric motors for most of the work, because these are very well known components, easy to characterize and control,” says Hammond, who arrived at Georgia Tech in time for last fall semester.
By contrast, he says, “soft robotics is a very nascent field. Controlling softer, compliant mechanisms is much harder.”
Hammond gets the appeal of soft robotics and soft sensing and wearable devices in general, adding, “but I also see the appeal of having extra robotic limbs or even mobile robots that can help us do things collaboratively. I’m trying to fuse all of that into my research efforts here at Georgia Tech.”