Utilization of machine learning methods for assembling, training and understanding autonomous robots

For decades human society has been supported by the proliferation of complex artifacts such as electronic appliances, personal vehicles and mass transportation systems, electrical and communications grids, and in the past few decades, Internet. In the very near future, robots will play increasingly important roles in our daily life. The increase in complexity of the tasks and sometimes physical forms or morphologies of the artifacts consequently requires complex assembling and controlling procedures of them, which soon will be unmanageable by the traditional manufacturing process. The aim of this paper is to give a brief review on the potentials of the non-traditional assembling of complex artifacts, which in this study is symbolized by the creation of autonomous robots. Methods in self-assembling modular robots, real time learning of autonomous robots and a method for giving the comprehensive understanding, albeit intuitively, to human will be explained through some physical experiments.