Robotics: Robots learn to walk and read
Rotober are more human: move safely in unfamiliar territory and can even learn to read like a man.
The EU project Local Morph examined the robot locomotion and walking robot tries to improve through efficiency, robustness and the ability to get around in unfamiliar surroundings. The focus is in two areas: the optimal shape and the ability to adapt through a change in shape, for example, then when a robot is damaged or a component fails.
By exploring these two central ideas, the researchers want to develop robots that have an optimum maneuverability, self-stabilization, energy efficiency and a high degree of adaptability to unknown environments. As soon robots will be built that will surely move around independently of the surface.
Biologists, physicists and engineers from Belgium, Canada, Denmark, Germany and Switzerland to work at the optimum morphological factors, such as shape, material, sensor technology and body segmentation. They are currently studying different animals to decrypt their locomotion strategy and incorporate the results into robotics.
amputee learn from dogs
Researchers at the University of Jena studied the strategies are dogs that one leg had to be amputated, adapt to the new situation. They let the animals walk on a treadmill and recorded their movements with highly specialized infrared cameras. Since dogs are designed to travel on four legs amputated animals must organize their entire transportation system. The researchers found that dogs with an amputation of the hind leg can easily adapt to the new situation, such as dogs, which was missing a front leg. Then a three-dimensional computer model of the video data, which reflects the movements in detail. The results should help to give walking robots appropriate strategies to adapt when a component failure or damage from the outside, to the new situation.
Orient as a man
At the University of Applied Sciences, a robot has learned to read as a man. This allows them exactly how a person based on signs, knows where he is and the man can control it on the basis of human information. Outdoor robots are guided by GPS – in closed buildings, this often does not. Therefore, the student Christoph shepherd had the idea that a robot should be based on the basis of door signs in the building. This is achieved by the robot using a high resolution digital camera with more optical zoom in four steps:
1. Detecting a door plate
2, Locating door signs in a hallway
3. Photograph of the door plate
4. Text recognition using a standard OCR program
So that the robot is reading like before, like man, see letters, recognize letters, put letters together to form words, derive meaning and finally apply content.