Flying robots. Crawling robots. Swarming robots. Soft robots. Wearable robots. All of these robotic types are in development at the Harvard-based Wyss Institute for Biologically Inspired Engineering.
The Wyss (pronounced “veese”) Institute was established in 2009 from a $125 million gift from Swiss entrepreneur Hansjorg Wyss, who earned a Harvard M.B.A. in 1965. At the time, the Wyss donation was the largest gift Harvard had ever received. Wyss later doubled his donation to $250 million.
The mission of the Wyss Institute is “to discover the engineering principles that Nature uses to build living things, and to harness these insights to create biologically inspired materials and devices that will revolutionize healthcare and create a more sustainable world.”
The Wyss Institute is housed at two facilities: in the Center for Life Science Building in the Longwood Medical Area of Boston and in a Harvard campus building near the Harvard museums in Cambridge.
The Wyss Bioinspired Robotics labs are located on the Harvard campus, and comprise one of six Wyss Enabling Technology Platforms, focused on the development of cutting-edge biologically-inspired technologies and capabilities.
Harvard post-doctoral researcher Zhi Ern Teoh must assemble the delicate robobees under a microscope. Weighing a mere 80 milligrams, the robobee’s body is made from a carbon-fiber laminate, the wings from ultra-thin plastic. The wings are controlled by a minuscule flight muscle or actuator that drives wing movement when a voltage is applied through a thin strand of wire (see video).
The wire has a tethering effect, preventing the robot from flying freely. But it is not practical to power the robobees using an onboard battery, which weighs too much. Other power alternatives are being explored, said Zhi, including onboard solar collectors.
When Zhi, 31, tells his family and friends what he is working on, “They always ask whether these drones are going to used in reconnaissance, but we’re many, many, many years away from that.” At this point, he says, the robotic insect research is basic research, focused on giving scientists a better understanding of insect flight.
Zhi can foresee the day when there will be practical applications for the robotic insects. “You could potentially send a swarm of them into a collapsed building to search for people… The challenge of finding survivors in this unstructured environment is that wheeled and crawling robots may not be able to get into some of the really small holes,” he said.
But before the robobees can fly autonomously, there must be a number of technological advances, developing light-weight onboard control systems, sensors, and a power source. As Zhi said, inventing such miniature, advanced technology may take many years.
Cockroach-Inspired Crawling Robots
Fourth-year Harvard PhD student Ben Goldberg, 25, works on the Harvard Ambulatory Micro Robot (HAMR) team. The HAMR is cockroach-inspired, weighing the same as a cockroach (1.5 grams), with very similar leg motions to a real cockroach, said Goldberg (see video). The legs are controlled and powered by a thin tether wire connected to a computer. There is also an untethered version of the HAMR, which is controlled by an onboard microprocessor and powered by a tiny battery, which can supply only three or four minutes of power.
Other ways of powering the HAMR without a tether wire are being investigated, said Goldberg, including solar power, “which is not limited to the energy you can carry onboard, because it’s coming from an external source, the sun or a bright light.” The use of wireless electricity, similar to the charging pad technology used for cellphones, is also a possibility, he said.
The body of the HAMR is made from a Wyss-invented laminate, “carbon fiber, a flexible polymer called Kapton, and a dry-film adhesive. In total, we have 11 different layers. We laser micro-machine those materials…and we use this origami-inspired assembly method, where linkages fold into place,” said Goldberg.
The origami-inspired manufacturing technique developed by Wyss researchers is called pop-up MEMS (microelectromechanical systems). Post-doctoral researcher Jesung Koh, 33, has helped to perfect the technique. “My work is simplification of two-dimensional pattern design. Pop-up MEMS. The great thing about that is scalability. We can make small things and big things using the same pattern design… The main goal is super-cheap super-easy robots to make. Robots for everybody,” said Jesung.
Like the flying robots, the HAMR lab is engaged in basic research, trying to better understand the movement of crawling insects. “There’s no immediate application for that, aside from understanding how ants are moving… You never know what could come out of it. Just understanding the basic principles of how the feet are interacting with the ground when the robot is running” is important, said Goldberg.
When Swiss soft robotics expert Daniel Vogt, 33, was recruited four years ago by Wyss microrobotics lab director Rob Wood, he saw it as a “big chance to work with fabulous people on fabulous projects,” and accepted the invitation. He describes soft robotics as “robots which are made of soft materials, which are more suitable to manipulate delicate and deformable objects. It’s a new trend in research in robotics.”
Vogt describes a major advantage of soft robotics over hard. “If you have a hard-bodied robot and you want to manipulate an object, you have to know exactly where the object is and what is the size of the object. The power of soft robotics is that you don’t need to know the exact information. You can just approach the object and it will automatically conform to the object, whatever the shape of the object is.”
Vogt sees immediate applications for soft robotics. “If you want to carry an egg, for example, it will delicately surround the egg to hold it…whereas a hard-body robot will have very limited points of contact and it will be easy to put too much pressure and break the egg.”
There are environments where a soft robot, which typically has an exterior made of a silicone elastomer, would be better suited to carry out tasks than a metal robot, said Vogt, including tasks underwater or in acidic settings.
Nicholas Bartlett is a third-year Harvard PhD student and also a researcher in the soft robotics lab. Bartlett has a particular interest in designing the control system for the soft robots. “Control is traditionally done with electrical circuits, wires and sensors… For a number of reasons, that may not be ideal for soft robotics. For instance, a circuit board is typically rigid, so if you have a rigid part in a soft robot, it to some degree defeats the purpose of the soft robot. I’m looking at making a soft control system for robots.”
Bartlett says that considerable progress on soft control systems has already been made by Biomimetic Microsystems researchers at Wyss, who have developed a suite of medical devices called human “organs-on-chips”(see video).
Bartlett, 26, is aware that, when he tells people he is working on robots, they envision “a Terminator, a humanoid metal thing with joints.” But, he says, what people should think about is the octopus. “An octopus is typically an example that soft roboticists love to point to, something that’s bio-inspired and totally flexible. The inspiration is, how can we make our robots look more like an octopus, that can bend and move and squish itself.”
Soft robotics researcher Vogt commented on the uniqueness of the Wyss research environment. “The beauty of this place is, it’s like a Disneyland for researchers. We have a lot of fun…and we can very freely explore new spaces, new things. We’re not constrained, we don’t have any limitations, other than, it must be noble for science…Often it’s crazy ideas that lead to interesting projects. But that’s something that’s really encouraged here.”
Wyss senior research scientist Justin Werfel is an expert on complex systems, which he describes as “systems of a large number of independent components with interesting collective behavior.” His research work at Wyss includes social insect behavioral studies, and it includes swarm robotics, large numbers of simple limited robots, but with some powerful collective behavior. “Rather than robots that are inspired by humans, C3PO-style, it’s robots that are inspired by ant colonies,” he said.
Werfel, 38, said he fascinated by the ability of insects, such as bees, ants, and termites, to build complex nests and mounds by acting collectively. “For me, understanding that mystery is really the driving goal… The termite mounds, you’ve got millions of independent agents building large-scale complicated things. How does that happen? That’s quite a complicated system. When we first started, we used to say, ‘Oh, termites are so simple.’ We do not say that anymore. They’re not simple. They’re really remarkable things.”
To better understand how complexity emerges from innumerable discrete social interactions, researchers in the Self-Organizing Systems Research Group (SSR) built a complex robotic environment populated by over 1,000 small kilobots. As described at the SSR website, each kilobot moves independently, using a cellphone vibrator motor, and each is programmed with the same algorithms and the same task, which might be, “form the letter K” (see video). Over a matter of several hours, the kilobots carry out the assigned task.
“The value of the kilobots system is that it provides a platform for doing physical experiments on artificial swarms and seeing unexpected things that come out of that,” said Werfel. “Being able to build systems that have desired collective effects out of independent components is a way of showing that we understand a thing.”
Field research is also critical to understanding insect collective behavior, he said. Werfel has made three trips to Namibia, where large termite mounds can be studied. Working under a five-year NIH grant, he said, “The long-term goal is to learn about how the collective colony outcomes emerge from the individual termite behaviors.”
Conor Walsh, Director of the Harvard Biodesign Lab and a Wyss Institute core faculty member, says that, “when people think wearable robots, they think Ironman, they think of some big rigid exoskeleton.” But, the soft exosuit that Walsh, 34, and his team is developing is “very small, very unobtrusive, very lightweight, very low power.”
The goal of the DARPA-funded soft exosuit project is “to unburden soldiers. When soldiers carry very heavy weights, 45 kilos oftentimes, that really stresses their muscles and soft tissue. It leads to injuries. They get fatigued,” said Walsh. “The exosuit is giving bursts of energy at key moments during the walking cycle to be able to have a person’s muscles do less work, so that they can walk with increased efficiency” (see video).
But the soft exosuit research is not just aimed at assisting healthy people to perform more efficiently. For patients who have a gait deficit, often due to a stroke, the exosuit can help “to restore the function of their joints…to help restore symmetry and increase their walking speed and reduce their energy cost to walk,” said Walsh (see video).
Walsh contrasts the medical applications of the soft exosuit with a rigid exoskeleton. Patients who are paralyzed from the waist down have zero capacity to walk. A rigid, heavy exoskeleton can force such patients to walk, although in a slow, mechanical way. “The exosuit wouldn’t work for someone who is totally paralyzed. But there’s a lot of people out there who can walk, but can’t walk very well. The exosuit gives them the boost they need to help them walk more normally,” he said.
It may take several years before the soft exosuit is available to patients. “We’re seeking [commercial] partners who will be able to bring this to market. Should we find the right partner, you’re looking at the next two to three years to refine the technology a little bit more, run a clinical trial to prove its benefit, get it approved by the FDA, and have it available for sale,” said Walsh.
Another major research project underway at the Biodesign Lab is the soft robotic glove, which is intended to restore partial hand control to patients who have ALS, muscular dystrophy, or an incomplete spinal cord injury. Soft actuators behind each finger within the robotic glove allow the wearer to grasp objects firmly, thus achieving some degree of self-reliance and independence (see video).
Walsh says that the soft robotic glove has achieved proof-of-concept status, demonstrating that patients who wear the glove can realize functional improvement by restoring some hand control. Going forward, Walsh and his team plan to “take a fresh look at the engineering aspects…to reduce the form factor, reduce the weight, increase comfort, make it easier to put on.”
Patients who are unable to use their arms and hands often have small, atrophied muscles. But Walsh’s team has shown that the faint residual electrical signals in the arms, which can be detected using EMG, or surface electromyography, may be enough to control the gripping motion of the robotic glove. Glove control may also be possible using voice commands. Giving patients precise glove control will “end up being a combination of solutions,” customized to each patient’s abilities, said Walsh.
The Biodesign Lab team is moving ahead to make the soft robotic glove a real product. In the first half of 2016, they will be conducting a 10-patient trial of the glove, which will hopefully produce the data needed to approach a commercial company and to discuss bringing the glove to market, said Walsh.
Walsh praises the Wyss Institute for being a translationally-focused research institute, that is, for taking cutting-edge beneficial research breakthroughs out of the lab and translating them into real-world useful products. “For me, it’s very exciting as a faculty member, because I really want the work from my lab to have impact in the real world…. There’s a lot of people who just need a little bit of help in order to have significantly improved quality of life or physical performance,” he said. “When we’re working on a lot of these healthcare-related robotics technologies, the real impact comes when you can you actually have people use these in society and be able to benefit from them. The Wyss is a super-important piece of the puzzle in trying to do that.”