A team of researchers at UC Berkeley has developed an artificial intelligence algorithm modeled on human infants, allowing four-legged robots to quickly adapt to their environment.
Rapid engine adaptation, or RMA, allows robots to perceive the surrounding area through sensory joints in their legs. RMA allows robots to adapt to unpredictable environments and situations in seconds, says Deepak Patak, an associate professor of computer science at Carnegie Mellon University and co-author of the study.
“This adaptation model takes into account the history of the robot, as well as the force that walks on your feet,” Patak said. “When you walk on the sand, your feet are immersed in the sand. You can know it is sand by watching the force of your own foot as it sinks into the sand. ”
Previous efforts to improve robot mobility have typically involved reprogramming robots to adapt their gait to a wide range of deployed areas, Patak said. However, the problems arose when the robots encountered environments for which they were not designed to respond.
University researchers set out to solve the problem a year and a half ago, inspired by how children learn to walk and adapt to their environment during the first few years of life, says Ashish Kumar, a graduate student who co-authored the University project.
“We are inspired by how children learn,” said ite itendra Malik, a professor in the Department of Electrical Engineering and Computer Science and co-author of the program. “What is the secret?” The secret is that they watch the world, they interact with the world, they experiment with the world. ”
Researchers plan to further enhance RMA by adding visual sensors, extending the RMA to the robots’ fingers so that they can grasp objects, and even opening bottle caps.
According to Malik, RMA will allow these advanced robots to play a vital role in the care of the elderly.
“They need items brought, they need to be brought with them. If someone falls, you need 911 calls,” Malik said. “Animals serve some of these roles, but pets can’t open the fridge, take something out of the fridge while I imagine this robot does.”
Patak noted that the biggest scientific contribution to their work is the element of adaptability.
Applications of their adaptive model can extend widespread robotics to areas, including intelligence infrastructure. Artificially programmed ramps can facilitate traffic.
“Adaptation is a feature of intelligence,” Malik said. “There are different ways to adapt to change, but what we are adding is a powerful new technique.”
Contact Rachel Reps
قد يهمك أيضاً :-
- Clinical guidelines for children who are positive about SCID
- Cases of harming children are on the rise (և children are at the highest risk)
- Newborns beat the best investors to sell stocks, the study shows
- What about the children crossing the fence at Kabul Airport?
- Like human children, bats learn to communicate by barking
- The children of the Southland 2020 blockade are now happy little ones
- Sean Johnson Onson East Says He May Have Children