December 1, 2007 Computational neuroscientists designed a robot that teaches itself how to walk on differing terrain. The RunBot uses an infrared eye to detect slopes, and adjusts its gait for the smoothest walk. From repeated attempts it matches the degree of slope to the appropriate length of step and movement of joints. Essentially, the robot learns to walk on uneven ground in the same way as a child.
Robots are the wave of the future. And nothing is moving there faster than the world's fastest robot -- one that could set the pace for all robots in the future!
But what sets this robot apart from all others? It walks like a human and when it falls … it learns from its mistakes, and corrects itself.
“After two or three times, like a baby, it just learns from mistake and it changes gait and walks up the slope,” Poramate Mannonpong, Ph.D., Computational Neuroscientist at the University of Goettingen in Goettingen, Germany said.
Dr. Manoonpong is the computational neuroscientist behind the Runbot. By using sensors and an infrared eye, this robot can detect a slope on its path and adjust on the spot! It’s created with the same critical joints as a human.
“It has two hip joints, two knee joints, here … and it has a foot contact sensor for each leg.” Dr. Mannonpong said.
Just like in people, sensors make sure the joints are not over-stretched and that the next step is initiated as soon as the foot touches the ground. This technology may someday be used to help humans.
“We hope that in one day, human can walk with artificial leg,” Dr. Mannonpong said.
Speed walking robots started out slowly, at 0.7 leg lengths per second. Now, the Runbot can speed walk at 3.5 leg lengths per second. Compared to its size, that’s almost as fast as a speed walking Olympian. Proving robots are moving us fast towards the future.
The next phase is to make the Runbot freestanding and then work with companies to produce the technology for humans.
BACKGROUND: Scientists at the University of Goettingen in Germany have built a robot that teaches itself how to walk like a human being: the computer inside learns how to adapt the robot's stride to different terrains and environments. Runbot holds the world record in speed walking for dynamic machines. Just like a human, it leans forward slightly and uses shorter steps. It can learn this behavior after only a few trial runs.
HOW IT WORKS: The Runbot has an infrared “eye” (a sensor) that can detect a slope on its path and adjust its stride accordingly. The robot’s ability to switch quickly from one gait to another is due to the organization of the movement control. On the more basic levels, movement is based on reflexes driven by sensors. Control circuits ensure that the joints are not overstretched or that the next step is started as soon as the foot touches the ground. When the gait needs to be adapted, higher centers of organization step in. At its first attempt to climb a slope, RunBot will fall over backwards, since it has not yet learned to use its visual to change its gait. But like children learning to walk, the robot learns from its failures, leading to a strengthening of the contact between the eye and movement control. Once these connections are established, step length and body posture can be controlled by the visual signal.
HOW ROBOTS WORK: Robots are manmade machines intended to replicate human and animal behavior. Robots are made of roughly the same components as human beings: a body structure with moveable joints; a muscle system outfitted with motors and actuators to move that body structure; a sensory system to collect information from the surrounding environment; a power source to activate the body; and a computer "brain" system to process sensory information and tell the muscles what to do. Roboticists can combine these basic elements with other technological innovations to create some very complex robotic systems.
ABOUT A.I.: Robots and computer networks are always evolving intelligent consciousness in popular science fiction. But while modern scientists have made great strides in building computers that can mimic logical thought, they still haven't cracked the code of human emotion and consciousness. There are two prevailing schools of thought on artificial intelligence. Proponents of "strong AI" consider that all human thought can be broken down into a set of mathematical operations. They expect that they will one day be able to replicate the human mind and create a robot capable of both thinking and feeling, with a sense of self -- the stuff of classic science fiction. Think of the robot Number Five from the 80s movie Short Circuit, who suddenly realized, frightened, that he could be "disassembled" by the scientists who made him. "Weak AI" proponents expect that human thought and emotion can only be simulated by computers. A computer might seem intelligent, but it is not aware of what it is doing, with no sense of self or consciousness.
The IEEE-USA contributed to the information contained in the TV portion of this report.
Editor's Note: This article is not intended to provide medical advice, diagnosis or treatment.