Cognition learning: Brain-wave for robotic grasping and dexterity enhancement

Dexterous manipulation by multi-fingered robotics hands has been a research topic for more than thirty years. However, forming grasping closure and determination of fingertips forces has always been an issue due to related complexity. Will a robotic hand grasping be enhanced via learning from human grasping? There have been a number of attempts for that. This article is presenting a different approach, i.e. using a cognitive learning method for generating grasping forces. This is based on using brainwaves signals as a cognitive support for aiding in relating true grasping forces. Respectively, instead of using conventional analytical approach for distributing fingertips forces (a cumbersome approach), we shall rely on learning the brainwaves patterns being used by human grasping. However, there are number of issues related to this approach, e.g. detection of brainwaves, massive data collection-reduction-classification, and patterns learning, are crucial issues to be tackled and resolved.