Abstract
These days, the Internet and mobile networks are actively utilized, and a lot of data such as moving images and voices is constantly exchanged on a world-wide scale. Such physical communication as a handshake and a hug that is very important for humans is expected to further be installed to indirect communication via machines or communication channels, but it is still limited to research. This report discusses a generic framework in which persons at different places sophisticatedly communicate physical information by machines. Note the following two approaches: First, we derive system dynamics by minimizing the total energy composed of mechanical energy and constraints necessary for desired motions. Second, we enhance our framework for a module-based dynamical neural network model. This report also argues that our top-down method is more effective than a conventional bottom-up method.
