Evolutionary robotics

Evolutionary roboticsThe Evolutionary Robotics (ER) is an area of autonomous robotics, in which drivers are developed through evolution robots using genetic algorithms.

It usually evolves Neural Networks as it follows the reference biological and simplifies the design and representation in genetic algorithm.

The foundation of the RE is relationship to work at the CNR in Rome in the ’90s, but the initial idea of encoding the control system of a robot in a genome and evolutionary data of the 80 late. Continue reading

Evolutionary Robotics (Evolutionary Robotics)

225362-286784This approach applies the knowledge gained from the natural sciences (biology and ethology) and Artificial Life (neural networks, evolutionary techniques and dynamical systems) on real robots, to develop their own skills in close interaction with the environment and without human intervention.
With a fixed design, it is difficult to have a robot suits (self-organized) to a dynamic environment that evolves through-often-chaotic changes. Hence, the evolutionary robotics can provide an adequate solution to this problem because the machine can automatically acquire new behaviors depending on the dynamic situations that occur in the environment where it is located.

Through the use of evolutionary techniques (genetic algorithms, genetic programming and evolutionary strategy), you may decide to evolve the control system or certain features of the robot body (morphology, sensors, actuators, etc.). Or co-evolve both. Similarly, you may decide to evolve physically the hardware (electronic circuits) or software (program or control rules). However, little is done on evolvable hardware [Fernández León, 2004] and usually what is done is to move first driver in a computer simulation, and only then, are transferred to real robots. The robot controller typically consists of artificial neural networks, and evolution is to change the weights of the connections of the network.

Currently, the main drawback is its slow evolutionary control convergence speed and the considerable amount of time that must pass to complete the evolutionary process on a real robot [Pratiharas, 2003]. It is not appropriate for solving problems of increasing complexity [Fernández León, 2004].
Robotics Biomimetics, Biologically Inspired Robotics Biorrobótica or
This approach is concerned with designing robots that function like biological systems, hence they are based on the natural sciences (biology, zoology and ethology) and robotics. Given that biological systems perform many complex processing tasks with maximum efficiency, provide a good reference for implementing artificial systems that perform tasks that living things do naturally (interpretation of sensory information, learning, movement, coordination, and so on. ) [Ros, et al, 2002]. Although it is possible to obtain different degrees of “biologically inspired” (from a vague resemblance to an acceptable reply), the ultimate goal is to make machines and systems increasingly similar to the original [Dario, 2005].

The advantage of building bio-robots is that, as is possible to study all their internal processes, they can be contrasted with the different organs of the animal from which it is based. Currently, scientists develop locusts, flies, dogs, fish, snakes and roaches robotics, in order to emulate a greater or more behavior-robust, flexible and adaptable animals. However, few machines resemble their natural counterparts.

Replicate biology is not easy and could be some time before they can produce biomimetic robots that are truly useful. Another problem, perhaps the most-is that, although well aware of the different processes of many of these living beings, there is a huge difference with their human counterparts. Indeed, the manner in which man perceives and acts is extremely more complex than a lobster does, to give an example.

Sergio Alejandro Moriello is Electronic Engineer (1989), Postgraduate Diploma in Science Journalism (1996), Postgraduate Diploma in Business Administration (1997), Specialist in Information Systems Engineering (2005) Studying Masters in Information Systems from UTN-FRBA ( Thesis completed). Author of books Intelligences Synthetic and Natural and Synthetic Intelligence.