Study enough, kids, and maybe one day you’ll grow up to become a professional fighting robot. A few years ago, Boston Dynamics set the standard for the field by having people using hockey sticks try to prevent the quadrupedal robot from opening a door. Earlier in 2015, the remote federal research agency Darpa hosted a challenge in which it forced the clumsy humanoid robots to be embarrassed in an obstacle course. way out of the league of machines. (I once asked you, dear readers, to stop laughing at them, but since then I have changed my mind.) And now, here it is: the creators of the robot dog Jueying have taught him a fascinating way to defend- if of a human antagonist who strikes him. envelope or push it with a stick.
A team of researchers from Zhejiang University of China, where Jueying hardware was also developed, and the University of Edinburgh not teach the Jueying how to recover after an assault, as much as they let the robot find out. It’s a spectacular deviation from how a hardware developer like Boston Dynamics teaches a robot to move, using decades of human experience to code hard, line by line, the way a robot is supposed to react to stimuli like foot.
But there has to be a better way. Imagine, if you will, a football team. Midfielders, strikers and the goalkeeper usually do things related to football, such as running and kicking, but each position has its own specialized skills that make it unique. The goalkeeper, for example, is the only person on the field who can catch the ball with his hands without screaming.
In traditional robot training methods, you will need to meticulously code all of these specialized behaviors. For example, how should actuators (motors that move the limbs of a robot) be coordinated for the machine to function like a midfielder? “The reality is that if you want to send a robot into nature to do a wide range of different tasks and missions, you need different skills, right?” says Edinburgh University robot Zhibin Li, corresponding author of a recent article in the journal Robotic Science describing the system.
Li and his colleagues began training the software that would guide a virtual version of the robot dog. They developed a learning architecture with eight algorithmic “experts” that would help the dog produce complex behaviors. For each of these, a deep neural network was used to train the robot’s computer model to achieve a particular skill, such as jogging or straightening if it fell on its back. If the virtual robot tested something that brought it closer to the target, it would get a digital reward. If he did something that wasn’t ideal, he would get a digital demerit. This is known as reinforcement learning. After many of these trial and error guided trials, the simulated robot would become an expert in a skill.
Compare this to the traditional way of coding a robot to do something as seemingly simple as climbing stairs.this actuator spins so much, this other actuator spins so much. “The approach to AI is very different in the sense it captures experience, that the robot has tested hundreds of thousands of times, or even millions of times, ”says Li. “Therefore, in the simulated environment I can create all possible scenarios. I can create different environments or different configurations. For example, the robot may start with a different attitude, such as lying on the ground, standing, falling, and so on. ”.