Army Explores Equipping Robots with Living Muscle Tissue
A scientist from the Army Research Lab gave a glimpse into its early work with biohybrid robotics.
Looking to pave the way for the production of nimble robots that can move more like living creatures than bulky androids, Army Research Laboratory scientists are embarking on fresh, high-risk studies in biohybrid robotics that could eventually fuse organic tissue with machines.
“This is wholly new to the lab, and the field itself is still relatively young. The publications associated with the first idea of successfully integrating muscle tissue or cells into some larger architecture to control motion with that same biological device didn't really start until after 2000—and really spooled up in the early 2010s. So it's very young, even as a discipline,” explained Dr. Dean Culver, a research scientist at the laboratory. “And that kind of gave us an opportunity to see how we could help move it forward and what expertise that we had that we could levy in that direction to really make an impact.”
In a recent interview, he briefed Nextgov on this future-facing research and the breathtaking applications it could spur.
A Potential Huge Step Forward
Culver studied mechanical engineering, and while in graduate school at Duke University, he became increasingly interested in energy management, and what he deemed the complicated ways beings get motion out of stored energy.
“After I graduated, one of the natural extensions of that is ‘Hey, how exactly do muscles work? How do organisms store energy and turn that into motion?’” he noted. “And it turns out that we knew less about the answer to that question than I had originally anticipated. So, there are obvious applications of that in robotics, and the design of mechanisms and new vehicles, for the Army. That brings us to today—I'm still working on that problem.”
It’s a complex pursuit that Culver has been tackling with the lab since about 2017. Ultimately, he envisions being able to give engineers the capability to design devices that last a long time, are really resilient, quiet—and don’t run hot.
“And those are all things that biological systems offer,” Culver explained.
Some existing, state-of-the-art robots that the Army uses can carry things across various locations, or record what their surrounding environments look like. They typically have four wheels, are a foot or two in height, and move over paved terrain at about two miles per hour. But there’s an issue, according to Culver, who noted that researchers are reaching a point where they’re experiencing diminishing returns in the design of these robots with wheels as their primary locomotor, and batteries as their centralized power system.
“We look at a wolf in nature: It probably weighs about the same, can pull much more and can travel hundreds of miles without really eating, take a nap and do the same thing the next day,” he said. “There's a huge performance disparity between those two things. And if we can offer the ability for robots to go out on these long missions, based on these design principles that we can understand from observing nature—that's a huge step forward.”
The idea is that instead of pushing this “battery-motor” paradigm forward, the research community could save heaps of time by potentially applying contractile actuators like muscles, and distributed chemical energy maintenance, instead of just batteries. Culver added that a real advantage of muscle tissue and the architecture that biology uses to turn energy into motion is that it's a bit flexible. For example, if he told today's top quadruped robot to run across a parking lot, it can do it pretty well. But, he noted, if he commanded it to run off of the parking lot onto a gravel surface—without telling it about the change in surface—“it's not going to have a good time.”
Part of that is strung from the rigidity of the machine’s actuator, as those motors aren't ready to accommodate something surprising.
“But if you run through a field, and your footsteps into a rabbit hole, even before the signal from your foot has reached your brain to say, ‘Oh, my gosh, I'm in a rabbit hole,’ your body is already moving to accommodate that sudden change. Part of that is the way that control systems are designed in organisms—that's obviously really amazing—but another part of that is the ability of muscles and tendons to bend and flex a little bit, and offer those control systems an opportunity to adapt. So that is a huge capability that we could offer,” Culver said. “Robots who are, obviously, in Army applications going to go into unknown and unpredictable environments—they need to be able to adapt to things that they weren't planning for. So, that’s a big part of this effort as well.”
ARL isn’t at the point where it’s produced full-blown biohybrid prototypes. Still, those could come. And as Culver put it, “the idea, really, is that if you can put the tissue connected to two solid surfaces—and you can turn on the muscle, turn it off with either electrical impulses or chemical actuation,” it would then contract, “sort of like a spring that you can tell when to shrink and expand, and there's some great advantages to that over motors.”
For such items, muscle tissue used from a variety of animal species could be grown in a lab, based on the Army’s needs.
As for his own present role in this highly technical landscape, Culver said he’s most interested in the math and theory behind controlling and regulating these elements—and pinpointing relevant use cases. Eventually, when the Army moves to design robots, he wants to be able to say whether a motor or biohybrid muscle actuator makes the most use for specific applications.
“In order to really answer that question, we need to have a really clear understanding of how that muscle tissue works and how it can be integrated into a system that gets it the electrical impulses, ions and chemicals that it needs to get its job done really well and lasts a long time,” Culver explained. “Right now, as I advance that forward, we're looking for academic partners that can run experiments to validate the models that I develop.”
Initial biohybrid robotics applications ARL’s team expects to go after involve “legged platforms similar to the Army’s Legged Locomotion and Movement Adaptation research platform, known as LLAMA, and the U.S. Marine Corps’ Legged Squad Support System, or LS3,” the lab confirmed in its release, noting that those participating are also considering flapping-wing drones.
So far, collaborators from Duke University and the University of North Carolina linked up with ARL for these unfolding experiments.
During the conversation with Nextgov, Culver emphasized the importance for military researchers to be “leaders in these disruptive areas.” Amid this realm, particularly, playing catchup with American adversaries could prove dangerous.
“Biohybrid and bio-inspired engineering, at this scale and breaking it down into a framework like this, it's more than saying, ‘OK, we're going to take what's in biology, and we're going to harness it.’ It's saying, ‘We're going to learn the principles that make biology great at what it does, and now we're going to strip away all of the things that hold biology back. We're going to learn how to build things from the ground up, and not only offer robots and devices and mechanisms, the capabilities that we see in nature—but just revolutionize the way that we think about it and push the horizons on what we're capable of way further back,’” Culver said. “And that disruption is precisely the kind of thing that we want to lead, and that [ARL] should be front-runners in.”