- Researchers at UC Berkeley are studying fox squirrels to understand why they are such precise jumpers.
- Squirrels have some sort of extraordinary judgment that allows them to perform risky jumps every time.
- In the future, this research could allow military robots to make split-second assessments and infinitesimal adjustments when navigating complicated terrain.
At the University of California at Berkeley, science is paving the way for future U.S. Army robots to roll or walk autonomously over rough terrain with ease, thanks to new research on an unlikely hero: the squirrel common.
Specifically, researchers are studying fox squirrels in an effort to understand why they are so good at making treacherous jumps to reach delicious peanut treats. In fact, squirrels are so good at “reading” obstacles with their own physical limitations in mind, according to the army, that they almost never fall.
This quality is the complete opposite of what most robots can boast. Locomotion is a complicated subfield of robotics for a reason: machines are good enough for publication date and get stuck. So, as we think through it, a bio-inspired approach – using the ubiquitous backyard squirrel as a test case – could help military and search-and-rescue robots negotiate even the most rugged terrain. .
While the U.S. military has deployed a wide variety of flying drones over the past two decades, from the original MQ-1 Predator to the airliner-sized MQ-4 Global Hawk, the development of drones in ground is late. Unlike flying drones, which rarely need to face obstacles or tricky situations, ground drones navigate a more complex environment. Ground drones can face a range of obstacles, some navigable and some not. Detecting if he can climb a pile of rocks or jump through a trench, for example, are the kind of tricky questions a robotic brain needs to be able to answer quickly in the field.
For decades, roboticists have leaned on other living creatures, such as the gecko and the cockroach, in their quest to build nimble robots capable of overcoming difficult situations. But now the focus has shifted to an even trickier problem: discovering how robots can learn to make locomotion decisions in a fraction of a second based on their own mechanical limitations.
Enter the squirrels. In a eucalyptus grove on the UC Berkeley campus, as part of a project funded by the U.S. military, the National Science Foundation and the National Institutes of Health, researchers studied wild rodents, manipulating them for them to jump for treats.
Squirrels possess the incredible ability to judge an obstacle with their own agility in mind. They can investigate a problem, such as reaching a peanut suspended above the ground, and determine if a course of action, such as jumping from a log to a treat, will be successful. Squirrels are not always perfect for judging a jump, but they never experience catastrophic failure. Although they can stop to assess and even make a few false starts, squirrels never actually fall while performing the jump. In the worst case, they can also use their claws to hold on tightly to trees.
Dean Culver, program director for Dynamics and Complex Systems at the U.S. Army’s Combat Capability Development Command, explains what scientists hope to learn.
“The battlefield, especially the battlefield of the future, is an unpredictable place,” he says. Popular mechanics in an email. “If we want robotic platforms to go wherever the Warfighter can and more, these platforms will require quick and creative decision-making at a low energy cost surrounding the tasks that humans take for granted. Like deciding whether to jump or go around an obstacle.
This work could also help autonomous robots make better decisions in real time, Culver says. “By studying how organisms (like the awesome squirrel) decide how to jump and land to achieve their goal, we can learn more about how future engineering systems make decisions and act to respond to scenarios on which they were not trained. “
Berkeley scientists are also studying how their work could benefit search and rescue robots. Robots that enter collapsed buildings, earth cracked ground by earthquake, or other disaster-stricken areas will encounter many obstacles in their path. “The same principles that apply to the necessary performance capabilities of a robot on the battlefield of the future apply to search and rescue robots,” Culver says, “except their task is specified. Environments are always unpredictable and these robotic agents will require the same adaptability. “
The battlefield – and the disaster area – is a dangerous place, strewn with obstacles. The robots of the army of the future may not look like squirrels, and they certainly won’t jump for peanuts, but their internal decision-making abilities may one day be modeled on them.
Now watch this
This content is created and maintained by a third party, and imported to this page to help users provide their email addresses. You may be able to find more information about this and other similar content on piano.io