[A model of bipedal walking adaptable to an unpredictably dynamic environment].

Modern science has been developed through concept of subject-object separation. That is, nature has been cordoned off from human beings and objectified. We have attempted to discover ideal world laws wherein we can consider nature as homogeneous. The real world, however, is by far more complicated than what natural sciences have so far been able to decipher. There are many problems that cannot be effectively addressed with the existing scientific technology. Because the real world is so unpredictable and dynamic, it is impossible to objectify it in advance and apply traditional methodology. This real world problem arises especially in information processing systems, that is, the recognition and the motion control systems coping with the real world. The current information systems can only handle explicit and complete information. Life is an intrinsic part of nature. To be both pliant and sturdy in a complex environment requires autonomy capable of creating the information needed to control the self. It forms the premise for the cognizance and control of life systems that exist in reality. To "live," a life system must independently forge a harmonious relationship with an unlimited environment. It requires that the life system be capable of creating the information necessary for self-control. It is this autonomy that clearly distinguishes the world of life systems from the physical world. Here, we will show an example of adaptive bipedal walking under an indefinite environment.