Cassie is the “Usain Bolt” of the robot world.
Cassie’s time is 24.73 seconds, slower than Usain Bolt’s record of 9.58 seconds, but she has no rival in the two-legged robot world, Oregon State University announced.
About 40 people came and supported Cassie on the podium. Everyone considers their success to be a great moment in the history of robotics. Washington Post. Cassie’s speed and agility were trained by artificial intelligence, showing that the two-legged robot can maneuver in real-world situations while maintaining balance, a problem that has caused much trouble for scientists in the past.
The development team’s first achievement was that by 2021, Cassie was able to complete a continuous straight walk in 53 minutes. It was the result of five years of mechanical engineers and researchers at Oregon State University and Agility Robotics leading the way. for more demanding models.
“This is the first major step in getting humanoid robots to do real work in the real world. Now we can make robots move powerfully around the world on two legs,” said Alan Fern, a professor of intelligence. According to him, the AI helped train Cassie.
For decades, scientists, entrepreneurs, and engineers have favored robots with human-like bipedal structures. In the 1960s, Japanese researchers created the basic prototype of a bipedal motor. For the past decade, engineers at MIT and Caltech have tried to do the same. Tesla CEO Elon Musk last week unveiled a two-legged humanoid robot called Optimus.
Researchers say that two-legged robots have always faced the big problem of losing their balance and falling. To solve this problem, Fern teamed up with Jonathan Hunt, a professor at Oregon State University and co-founder of Agility Robotics, to train a biped robot using machine learning and neural networks, which are algorithms that mimic the workings of the human brain. The research was funded in part by the National Science Foundation and the Defense Advanced Research Projects Agency (DARPA), the secret government organization behind innovations like the Internet.
Since 2017, the team has been teaching Cassie to walk and using algorithms to “reward” the robot when it moves appropriately. “Everything is inspired by Pavlov’s psychology,” says Fern. “Learn to anticipate the rewards and do the right thing.”
Once the team is successful in the lab, the next step is to teach them how to behave in a real-world environment where surfaces are uneven, friction can change, and the robot’s mass also changes. With success in this year’s 100m, the team moved on to the next step: shaping Cassie’s torso to give her a human appearance.
A fully humanoid robot based on Cassie’s leg design, it requires peripherals to recognize terrain and navigate its entire body over rough terrain. In addition, the robot must recognize the object as something that can be picked up and then be intelligent enough to do it the same way a human would. “For example, if a robot is asked to place a box in a room, it needs to know how to find and retrieve the box and then place it somewhere else,” Fenn said.
Nancy J. Cook, Professor of Human Systems Engineering at Arizona State University, notes that robots can do amazing things like run or kick a ball. The tricky part is creating machines that interact naturally with humans. “What they lack is a very complex perception. Interacting with people still requires a lot of algorithms that they don’t have.”
Cook also says it’s commendable that cyborgs like Cassie are at the forefront of the robotics industry. However, he wondered if there was a need to build machines that simply mimic what humans do. It would be more valuable to build robots that can do things humans can’t, he said.