An Interdisciplinary Incubator

Lab members display the ATL soft-body robot prototype.

By focusing on how animals move, a group of Tufts researchers are changing how we think about (and may one day build) robots.

Located a half mile from the Tufts Medford campus at 200 Boston Avenue, the Advanced Technology Laboratory acts as an incubator to bring researchers together and speed up the evolution of ideas. Engineers use biological principles to help design and build structures, which in turn give biologists better ways to explain what they observe. Biomimetics, or mimicking nature, specifically in the form of caterpillars, was just such an interdisciplinary problem, and Barry Trimmer saw it as a way to galvanize researchers across disciplines.

"We began to think about ways we might design robots the way animals move around," says Trimmer, professor of biology and the director of the Biomimetic Devices Laboratory at the Advanced Technology Laboratory, or ATL. "There aren't any machines-any equivalent robots available-that can achieve the same kind of mobility as animals. My dog is capable of running through the woods, swimming in the water, climbing over bushes, and doing things that no robot on the planet currently can do."

Form and Function
Unlike dogs and other vertebrates, caterpillars don't have bones associated with their muscles to provide a system of levers.

"Though we know lots about vertebrate muscles-just shelves of books, because scientists are vertebrates-we don't know a whole lot about caterpillars' muscles, almost nothing, you could put it in one hand," says William Woods, a postdoctoral associate in the Tufts biology department.

Yet caterpillars can crawl up walls, grasp narrow branches or stems with sticky Velcrolike feet, and rotate their bodies almost full circle as they sense their environment.

"We got very excited about how an invertebrate animal like a caterpillar or octopus is able to control these very fluid, flexible and complicated movements using a very small nervous system," says Trimmer, who has worked with the tobacco hornworm, Manduca sexta, for more than 20 years. Soft-bodied biomimetics is a monumental task, especially when considering that nature has had millions of years of evolution to test adaptations and try new ways to solve problems. Despite nature's head start, the ATL crucible creates a space for Tufts engineers and scientists to turn up the heat on generating ideas.

"Toward the beginning of this project, we realized the best way for this to work was to have all of the different disciplines working together in the same laboratory space," Trimmer says. "It rapidly became clear that this was a very difficult problem that needed expertise in all sorts of areas outside of biology."

"The origins of the Advanced Technology Laboratory came from a collection of faculty who were seeing more and more opportunities for interdisciplinary research," says David Kaplan, chair of the biomedical engineering department and one of the primary investigators at the ATL.

The push for a collaborative research space like the ATL came not only from interested faculty, but also from Tufts' university-wide commitment to interdisciplinary research and the School of Engineering's strategic vision to create such a facility. Help from Tufts University and a $730,000 grant from the W.M. Keck Foundation gave legs-albeit squishy ones-to the creation of the ATL complex and the Soft-Bodied Robots project. Understanding soft materials is one of the first steps to building a robot that, like a caterpillar, couldn't be made out of traditional hard materials.

"Nature does a much better job than we do in creating materials," says Luis Dorfmann, an associate professor in the civil and environmental engineering department and an expert in soft materials. "If you look at the materials we came up with, it's very impressive?steel, concrete?and it's used everywhere. But now look at nature. The materials which nature develops, they change, they transform, they grow, they age. If you cut your skin, it heals itself."

Though the soft-body robot prototypes won't be made with a material that can mimic all these properties, eventually the ATL group will make the softbots out of a biodegradable material, like silk.

"By developing a soft, flexible material that can degrade, a soft-bodied robot made of this material could enter the body as a diagnostic tool and not need to be retrieved," says Kaplan, who researches bioengineered, biodegradable silk structures.

Arts, sciences, and engineering faculty members and graduate students discuss ways to control the movements of the soft-body robot.

A Natural Model
Creating a biodegradable softbot caught the attention of the U.S. government, which recently awarded the ATL research group $3.3 million to continue work in this area. Dorfmann and his graduate students in the ATL's Mechanics of Soft Materials Lab are looking at a stretchy silicone polymer, called Dragon Skin?, used on movie sets to make fantastic animatronic masks and monsters.

Erica Belmont, a Tufts chemical engineering alumna and current mechanical engineering graduate student in Dorfmann's lab, translates motion of the caterpillar's stretchy skin into a mathematical model that could eventually be recreated using Dragon Skin or a similar, complex soft material like silk.

"Instead of a rough simulation, using hinged joints or using hard materials, we can use soft materials to create the actual motion of the caterpillar that currently can't be recreated," says Belmont.

"And if we run into problems with not knowing how to model something in the robot, we can go back and check how the animal does it," says Linnea Van Griethuijsen, a biology graduate student who studies caterpillar locomotion at the ATL.

First chilling the caterpillars on ice, Van Griethuijsen can then apply tiny markers to their legs. After the caterpillars warm up, she can track the markers with a computer as they crawl along different surfaces.

"They have a wave of steps starting from the posterior end," she says. "We use the computer to see how high they lift their legs, how fast, in which dimensions, and in what order. This might give us a template for how we'd like the robot to move."

Though the caterpillar locomotion is complex, the nervous system isn't. Michael Simon, a graduate student in Trimmer's lab with a background in engineering, researches how much information the caterpillar's simple sensory organs register. Van Griethuijsen's research gives Simon an overall picture of movement, but it doesn't tell him how the caterpillar understands where its legs are when crawling.

"How does an animal with so few neurons control such a complicated body?" asks Simon. "To find out, I can open up the animal and remove the sensory organ and play with it, move it around, pull on it as if it was in a crawl sequence and see what kind of activity that generates," says Simon. "What you find is that you don't need all the information in the world to make the animal move. All you need are just a few parameters."

Simons' work gives Valencia Joyner a better understanding of how the caterpillar controls its movement and how to mimic that in an electronic system.

"We're really trying to craft the brain of the robot," says Joyner, an assistant professor in electrical and computer engineering and head of the ATL's Advanced Integrated Circuits and Systems Laboratory. "We're looking at innovative ways that we can use nanoscale circuits and systems that can easily interface with these new soft, bionic materials."

By using off-the-shelf electronic components and building new microscale chips, Joyner can develop a set of nerve-like electronic signals to stimulate the biomimetic materials.

"We want to find the most power-efficient way to deliver patterns of signals to cause the biomimetic skin to scrunch and move the way the caterpillar scrunches."

Getting the power to the circuits to create these signals is another piece of the interdisciplinary puzzle being worked on by Sameer Sonkusale.

"Existing circuits use a lot of power, but our implants must essentially run on no battery," says Sonkusale, who is an assistant professor in the electrical and computer engineering department and head of the ATL's Nanoscale Integrated Sensors and Circuits Laboratory. If the softbot caterpillar was sent into a collapsed building or an area with lots of debris as part of a search and rescue "you want to transmit the images to a central base station saying ?This is what I see,'" adds Sonkusale. "And you would want to sustain that softbot's power for a long time."

One way to generate power might be energy harvesting from the environment. Metal coils (or cantilevers) inside the robot would receive power from an external electromagnetic power source wirelessly. Energy is transferred without a direct connection through what's called an inductive link. To create coils for inductive links within the soft materials, Sonkusale had a discussion with Gary Leisk, a research assistant professor and mechanical engineer at the ATL.

"First, we needed to know how to mesh the coils together into the fabric," Sonkusale says.

Leisk put one of his senior mechanical engineering students on the task.

"The things that I thought were very difficult to achieve mechanically, they just didn't feel that way. In six months, we had some really nice devices to work with and play with," says Sonkusale. "That's the beauty of interdisciplinary collaboration. When you limit yourself to your discipline, sometimes you can't see what you can achieve."

Take-Home Lesson
The experience of learning in a collaborative environment is especially key for students, says Trimmer.

"The environment at the ATL allows us to train students deeply in a discipline but be able to have them communicate across disciplines."

Linda Abriola, dean of the Tufts School of Engineering, also praises the interdisciplinary nature of the lab.

"The ATL is a model for how science and engineering should approach multidisciplinary problems?faculty and students from different departments and educational backgrounds are working side-by-side in the same research space," says Abriola. "This dynamic interaction between engineering and biology is producing innovative research, as well as a collaborative learning environment for students."

Belmont agrees. "My experience at the ATL has been great, from being introduced to everything from biology to electrical engineering, and really finding what I can bring to these new disciplines and what they can bring to what I'm working on," she says.

With information learned from his engineering counterparts, William Woods says he now sees his own work in biology with new eyes.

"I came in here looking at this problem as a muscle physiologist might," he says. "If there's been one whopper of an unexpected take-home lesson, it's that the mechanical properties of the muscle are critical."

"This project has made me start thinking about if we can make the products we work with every day out of soft materials. Would it be easier for you if your cell phone could mold to your ear? Or if your iPhone could be slapped against the wall or against the table and regain its shape without any damage?" Michael Simon asks. "It's like I've come full circle, in a way: learning about engineering, then learning about biology, and in the end, understanding how there's another way to think about engineering."

To learn more about the Tufts Advanced Technology Laboratory, call 617-627-0900.

Article by Julia C. Keller, communications specialist, Tufts University School of Engineering.

Photos by Melody Ko