Harvard's UrchinBot Is One of the Weirdest Looking Robots We've Ever Seen
On the spectrum of weird stuff that can be found in the ocean, sea urchins are probably somewhere in the middle. They're an interesting combination of rigid and flexible, with shells covered in hard movable spines as well as soft tubular appendages that work like a combination of legs and sticky feet. The mobility strategy of sea urchins leverages both of these appendages, and while they may not be speedy, they can get themselves into all kinds of potentially useful nooks and crannies, which seems like a capability that could be valuable in a robot.
At IROS 2019 this week, roboticists from Harvard presented a bioinspired robot that "incorporates anatomical features unique to sea urchins," actuated by pneumatics or hydraulics and operating without a tether. It may be based on a real animal, but even so, UrchinBot is definitely one of the weirdest looking robots we've ever seen.
As it turns out, adult sea urchins are complicated critters, and making a robotic version of one of them was asking a bit much. Juvenile sea urchins incorporate the same basic features in a much simpler body, and while they're only 0.5 millimeters in size, a scaled-up version (with a body 230 mm in diameter) was much more feasible.
Just like the adults, sea urchin babies have two mobility appendages: movable spines, and sticky tube feet. The physical resemblance is striking, but it's much more than just aesthetics, as the researchers emphasize that "particular attention was paid to accurately replicating the geometry and ranges of motion of the anatomical features on which our design was based."
UrchinBot's spines (which the real animal uses for protection, mobility, and to jam itself into crevices) reflect the two different kinds of spines that you see on juvenile urchins. Nobody's quite sure why the babies have fancier spines than the adults, but UrchinBot replicates that detail too. Each spine is connected to the body with a ball joint, and a triangle of three pneumatic domes around the joint can inflate to push the spine in different directions. All of the domes are interconnected inside of the robot which means both that the spines can't be actuated separately and that you get a satisfyingly symmetric rotational motion whenever the spines move. As they rotate against a surface that UrchinBot is resting on, the robot slowly turns itself in the opposite direction.
The tube feet are a little more complicated, because real urchins excrete sticky stuff that they use to glue themselves to surfaces, and then excrete an enzyme that dissolves the glue when they want to move. UrchinBot instead uses extendable and retractable toe magnets, which work perfectly well as long as the robot is moving on a ferrous surface. As the tube feet inflate, they move outward and angle their tips down, and with enough pressure, the toe magnets pop out and adhere. UrchinBot then reverses its hydraulics to suck the tube foot back in, pulling itself towards the adhesion point, and causing the magnet to pop off again once it gets there.
The rest of UrchinBot's body is taken up with pumps, valves, and electronics that allows it to operate completely untethered, both on land and underwater. Here it is in action:
It turns out that UrchinBot's spines exhibit a range of motion similar to that of an actual urchin, which is neat. The tube feet can achieve an extension ratio of 6:1, which is reasonably close to a juvenile urchin's 10:1 ratio, but much less than an adult urchin, which can extend its tube feet out to a 50:1 ratio. UrchinBot is not as fast as the real thing, which is to be expected with most bioinspired robots. Top speed is 6 mm/s, or 0.027 body-lengths per second, quite a bit slower than a juvenile urchin (which can hit 10 body-lengths per second going flat out) but only half as fast as an adult urchin.
UrchinBot may not be the speediest robot under the sea, but the researchers say that it could be useful for underwater cleaning and inspection applications, especially in situations where heavy fouling would be a challenge for more conventional robots. The priority for UrchinBot upgrades is to stuff it with as many extra actuators as it'll hold, with the goal of making the spines actuate individually and giving the tube feet extra degrees of freedom. While UrchinBot may not find near-term applications, it serves as a testbed to help researchers identify physical features and control techniques that could result in new types of more versatile and effective underwater robots.
"Design, Fabrication, and Characterization of an Untethered Amphibious Sea Urchin-Inspired Robot," by Thibaut Paschal, Michael A. Bell, Jakob Sperry, Satchel Sieniewicz, Robert J. Wood, and James C. Weaver from Harvard's Wyss Institute, was presented this week at IROS 2019 in Macau.