Researchers at Carnegie Mellon University are using the centuries-old concept of a telescope to develop new structures that could increase robots' flexibility and versatility in the future.
A telescoping structure is made of nested pieces which slide in and out of one another to different lengths. A classic, if outdated, example would be a pirate or sailor’s retractable telescope. Today, some ladders, umbrellas and tentpoles also use this technology.
Not coincidentally, these applications all share a common trait.
“Pretty much all of these telescoping structures move along just a straight line,” said Keenan Crane, Assistant Professor of Computer Science and Robotics at Carnegie Mellon University.
Crane, his colleague StelianCoros and Ph.D. student Christopher Yu had been considering these kinds of structures when they had a breakthrough.
“We had this kind of a-ha moment where we said, ‘Well, you don’t have to move just along a straight line. You can make a curve, you can make a twist, you can make all sorts of other shapes. And of course if you can do that, that leads into all sorts of new things beyond what we usually see," said Crane.
This realization led to the creation of unique software. Users can draw curves or import 3D shapes, which the program turns into telescoping pieces that can expand and contract smoothly. Structures can be simple, like a T-shape, or as detailed as a lizard or giraffe.
Within the program, Crane has simulated the movement of these designs: twisting, turning, collapsing, expanding and even rolling on top of wheels.
One design that has shown promise is a telescoping arm. In computer simulations, it can take almost any shape and bend in nearly any direction.
This kind of versatility in movement would augment the already considerable abilities of robotic arms, which are used in a variety of high-tech settings, fromoperating rooms to theInternational Space Station.
Crane said there are also potential non-robotic applications—he developed a model for tent supports, for example.
The ability to shrink down to a fraction of full length would also make any robot built this way easy to ship and potentially give it the ability tomove into smaller spaces. Robots could collapse inward to fit through a small opening, then extend back out on the other side.
Crane has already 3D-printed miniature versions of a few of the designs from the program. However, before any of these devices make the leap into robots, Crane and his colleagues need to figure out the best way to power and control their movements.
“I’ll tell you in a year,” said Crane, with a chuckle.
