Here comes the robotic brain!

Abstract

Year after year computers increase their processing power. This has been an advantage for the implementation of new and powerful robotic algorithms. While every day more complex algorithms appear, most of the increased processing capabilities of the newest microprocessors get easily exhausted. By analyzing the Artificial Neuron Model, this paper predicts that with the current tendencies in the increase of the microprocessor's power in 2050 the processing power of computers could reach the processing power of the human brain. Besides that, while the robotic systems integration becomes harder and no standards exist in this matter, the use of Artificial Neural Networks becomes more popular. This work studies the possibility of having a Neural Processing Architecture in the future, i.e. a Robotic Brain, under the perspective of pure neural processing power, as a feasible alternative in the development and integration of more complex architectures and algorithms in robotics. © 2010 Springer-Verlag.