Evan LernerNew GRASP Project Aims to Leverage ‘Embodied Intelligence’ via a Robotic Squirrel
It takes about a year before human infants master their own motor skills well enough to walk. While putting one foot in front of the other seems natural, remaining upright requires subtle shifts of balance throughout the body. Uneven terrain presents an additional challenge, but it’s one that children quickly overcome without much in the way of formal training or guidance. And once they’re up and running, there is no end to the novelty of jumping, skipping and climbing skills that kids discover and invent.
Legged robots don’t have it so easy. Only the most advanced can walk with a smooth, natural gait, and even those can be stymied by a small pile of rubble or sand. They have no capacity for inventing novel behavior at all: each new gait or maneuver must be programmed from scratch. And both humans and robots are put to shame by the average cat, squirrel or gecko, all of which can quickly adapt complex bodily features with manipulations to traverse almost any obstacle, whether familiar or completely novel.
A team of researchers, led from Penn Engineering’s GRASP Lab, now aims to imbue robots with this kind of “embodied intelligence,” developing bio-inspired designs that use limbs as sensors as well as actuators and learn new forms of locomotion based on interactions with their environment. Their five-year goal: a parkouring mechanical “squirrel” that will serve as a new paradigm for robot design and behavior.
The project is supported by the Army Research Office through the Multidisciplinary University Research Initiative, or MURI, Program funded by the Office of the Under Secretary of Defense for Research and Engineering. It is being led by Daniel E. Koditschek, Alfred Fitler Moore Professor in Electrical and Systems Engineering and interim director of Penn Engineering’s General Robotics, Automation, Sensing and Perception, or GRASP, Lab.
He will collaborate with Shu Yang, professor in the Department of Materials Science and Engineering, as well as Yuliy Baryshnikov of the University of Illinois Urbana-Champaign, Noah J. Cowan and James J. Knierim of Johns Hopkins University, and Robert J. Full and Lucia F. Jacobs of the University of California at Berkeley.
Animals’ bodies are exquisitely adapted to their habitats. Evolutionary pressure has resulted in bones, muscles and skin that automatically “solve” physical problems of force, sensation and energetics, allowing their brains to handle more abstract problems of placement, motivation, strategy and novelty.
“Animals are constantly performing both mechanical and mental processing of the information and energy flows at work in their environment,” Koditschek says. “‘Embodied intelligence refers to the integrated physical problem-solving and innovative capabilities that emerge from this interplay.”
For example, squirrels are at home whether on the ground, traversing tree trunks or leaping from branch to branch. They use their bodies expertly to solve different physical computations in each new environment. On level ground, their bounding gaits can be stabilized in part by changing their body’s mass distribution. Running down a tree engages mechanisms in squirrels’ ankles that bias the claws to “find” appropriately located grips. A leap is executed in part by tuning leg muscles to the springiness of the branch in the selection of a trajectory. Moreover, squirrels are constantly inventive, deploying suitably modified versions of previously discovered maneuvers as the need arises.
Engineers barely know how to even formulate, much less achieve such synergy in robot designs and programs. Koditschek’s own Kod*Lab has previously developed a series of agile, bioinspired robots that use their limbs as sensors. The lab has produced a robotic kangaroo rat, Jerboa, and a spin-off company, Ghost Robotics, which makes Minitaur, a quadrupedal robot that can open doors with a back handspring, among other tricks. These robots’ ability to “feel” the world with their limbs gives them uncommonly effective movement over a variety of terrains and the ability to manipulate objects encountered within them.
However, to achieve mobility approaching that of a real animal, like a squirrel, the researchers will work in three connected areas: the math that describes the information and energy exchange between the robot and its environment; how spatial information is encoded acted upon in biological brains and bodies working together; and new forms of robot body design, material construction and behavioral control that can take better advantage of both what the animals reveal and the mathematics explains about fine-grained sensing and control.
In a split second, sensations in the squirrel’s limbs need to be transmitted to its brain, interpreted by an internal accounting of the body’s engagement with the environment, and new instructions need to be sent back to the limbs to execute the correct movements. Long before this can happen, the material properties of the animal’s body have already begun to respond intelligently to the environment’s mechanical stimulus. Over a much longer time scale, the animal’s internal representation of where it is and what it needs to do relative to where it wants to be is adjusted as well. One of the MURI’s central aims will be to establish a formal understanding of the information and energy exchange at these different timescales between the environment and the animal’s integrated body-brain intelligence and to apply that insight to the construction of far more capable robots than exist today.
The MURI team features experts in animal cognition and biomechanics as well as neurophysiology and neuromotor control, promoting integrated study of the way the brain and nervous system interact with the muscles and skeleton in accommodating and influencing the environment.
Animals also have the advantage of highly articulated limbs, which allow for subtle sensing, balancing and grasping behaviors as they interact with the world. With this in mind, the MURI team includes experts in programmable materials who can design novel, kirigami-based, shape-shifting structures that will be incorporated into the robotic squirrel’s limbs. The group’s mathematicians will help translate between the body-brain structures that allow animals to place themselves in space, and the engineers’ active materials and mechanisms that drive the robot bodies.
“This is an absolute dream team,” Koditschek says, “focused on some of the most compelling questions about how animals work and capable of capitalizing on the answers to move robotics well beyond our clumsy present toward an agile and physically clever next step.”