8.2.10

darwin was right (again)


Inspired by the Darwinian principle of natural selection, evolutionary robotics views robots as autonomous artificial organisms that have the ability to develop their own skills in close interaction with the environment and without human intervention. Alan Turing suggested in the 1950's that intelligent machines capable of adaptation and learning would be too difficult to conceive by a mere mortals, but could be obtained through an evolutionary process with mutations and selective reproduction.

The general idea behind evolutionary robotics is to create a population with different genomes, each defining parameters of the control system of a robot or of its morphology. An initial random population of artificial chromosomes, each encoding the control system of a robot, is put into the environment. Each robot is then free to act (navigate, look around, manipulate) according to its genetically specified controller while its performance on various tasks is automatically evaluated. The fittest robots then reproduce by swapping parts of their genetic material with small random mutations. The process is repeated until the birth of a robot that satisfies the performance criteria. At the laboratory for intelligent systems, it was demonstrated that a couple hundred generations of random mutations and selective reproduction were sufficient to promote the evolution of efficient behaviours in a wide range of environmental conditions. The ability of robots to orientate, escape predators, and even cooperate is particularly remarkable given that they had deliberately simple genotypes directly mapped into the connection weights of neural networks comprising only a few dozen neurons.

SO CLOSE I CAN HARDLY STAND IT

1 comment:

  1. When they demonstrated this, was it with a math model? Or with actual robots with fake chromosomes? That would be awesome.

    ReplyDelete