Learning Of Robots
1st February 2010
According to the Swiss researchers, if the robots are allowed to evolve through natural selection then they will be able to develop adaptive abilities, for the purpose of hunting preys, cooperate and even they can help one another. Dario Floreano of the Ecole Polytechnique Federale de Lausanne and Laurent Keller of the University of Lausanne, have reported in a series of experiments described in the journal PloS Biology, that simple, small-wheeled Khepera and Alice robots can evolve behaviors such as collision-free movement and homing techniques, in only several hundred generations.
Neural network that mutated randomly has been controlling the robots, with input information from the sensors of the robots. In an imitation of natural selection, the robots with the best manipulating abilities were allowed to promote a new generation. Furthermore, having their neural net connections mixed and passed to a new generation paired selected robots.
It is being considered that within 100 generations, the robots would start moving through mazes, without even bumping into anything. Robots were found to cooperate in pushing larger tokens together to earn points for their whole group. The researchers have also observed altruistic behavior, in which robots would sacrifice getting points for themselves in favor of getting points for their whole group of related robots.
$100 Millions IPO Filed By Tesla Motors
Wind And Solar Feeling Heat From Low Natural Gas Prices The program also enable...











Download Free trial