
Class of ’94 | Drop by his busy laboratory on a typical weekday afternoon, and the odds are high that you’ll find Ralph Etienne-Cummings Gr’94 engaged in an esoteric research project—such as analyzing the electrical output produced by the swimming motions of a lamprey eel.
Or maybe you’ll catch him studying a housefly’s many-faceted eyeball in an effort to understand how the nerve-signals that help the insect avoid a fly swatter can be duplicated on a silicon microchip and then implanted in one of his own “seeing” robots.
A pioneer in the field of “neuromorphic engineering,” Etienne-Cummings directs labs dedicated to this rapidly evolving science at both the University of Maryland and the Johns Hopkins University in Baltimore, where he also serves as an associate professor in the Department of Electrical and Computer Engineering.
“At first glance, what I do looks pretty complicated,” says the 37-year-old engineer and roboticist, “but the purpose behind the research is actually quite simple. One of our key goals is to build electrical systems that will mimic the actions of living organisms, in order to make robots perform better by equipping them with biologically inspired sensory tools. Another long-term payoff for this kind of research will be electrical systems that can do the work of damaged nerve components in people with spinal-cord injuries, such as [the late actor and research advocate] Christopher Reeve.”
An electrical engineer who concentrated on developing “neural computers that could model human vision” while earning his doctorate at Penn, Etienne-Cummings often teams up with biophysicists on projects that combine data from biological and electrical systems in order to build microchips that can coax living structures (such as an eel’s spine) to move on command.
The lamprey project offers a good example of how research scientists are increasingly linking knowledge gained from biology with state-of-the-art electrical engineering. “We’ve been working with the lampreys for several years,” notes Etienne-Cummings, a native of Seychelles who studied in England before coming to the U.S. at age 18. “The beautiful thing about lamprey eels is that their spines each contain about a hundred or so bundles of ‘controller nerves’ that command the swimming muscles. And these ‘central pattern generators,’ as we call them, are what keep the creature moving steadily through the water.”
By electronically modeling the signaling that takes place in the eel’s pattern generators, Etienne-Cummings and his fellow researchers are gradually learning how to imprint the complex circuitry required for lamprey-swimming onto silicon-based microchips. “We’re at least 10 years away from being able to do the same thing with human spinal cords, which are much more complex,” he says. “Still, our work with the eels seems quite promising. We’ve nearly reached the point where we can send electrical signals to an eel-spine and get it to turn right or left on command.
“Eventually, we’ll be able to direct the swimming motions of an eel in a tank with our signaling boxes,” he predicts. “After that, we hope to move up the [mammalian] ladder to lab rats, cats, monkeys, and then to human spinal-cord tissue. It’s still a long way off, but we’re convinced that this kind of technology will one day help people who have been paralyzed by accidents or disease to regain the use of their limbs.”
By blending their knowledge of biology with state-of-the-art robotics, Etienne-Cummings and his colleagues are increasingly able to design tools that can help patients struggling with deficits caused by breakdowns in human nerve signaling. Not long ago, for example, he and a few of his grad students came up with a hand-held device that “sees” objects in the path of a blind person and then makes verbal suggestions on how to avoid the impediments.
“The science of robotics has really been taking off in recent years,” notes Etienne-Cummings, several of whose recent projects have been funded by the National Science Foundation and the U.S. Department of Defense. “One of the newest and most exciting trends is the development of robots that can actually understand their environment and then act on that understanding to modify their behavior.
“Increasingly, we’re seeing robot-machines that possess ‘situational awareness,’ which means that for the first time, really, they can effectively integrate sensory capability with action,” he explains. “These devices can adjust to their environment in real time—a step that will eventually lead to autonomous robots that can ‘think’ on their own and make decisions for themselves.”
While growing up in the “island paradise” of Seychelles—a group of tropical Indian Ocean islands near the east coast of Africa—Etienne-Cummings discovered that he loved doing math problems and building radios and antennas as well as “very dumb” robots, he says. His passion for working at the interface between biology and robotics “really took off at Penn” in the early 1990s, as he worked night and day to build microchips and computer programs that could teach computers how to see the world around them and then process the visual information.
“I had a blast working in the engineering labs, and I really came to enjoy the life of the graduate student,” he says. “I can’t tell you how many times we stayed up all night, working on our experiments and our computer programs. I remember fondly how we used to run downstairs at night and hurry over to the front doors of [HUP] to meet the food truck that always stopped by to sell hoagies and hamburgers and other goodies. That became a very important ritual after a while, and most of us wound up making the pilgrimage every single night.”
Ten years after leaving Philadelphia for Johns Hopkins, Etienne-Cummings and a few colleagues are close to perfecting a series of “robotic assistants” designed to help surgeons perform routine—and time-consuming—chores in the operating room. He predicts a great future for the field: “We’re going to be getting a lot more help from robotic technology in the years immediately up ahead.”
—Tom Nugent