A Modular Design Method of Machine Systems by Entropy Evaluation.

Flexible machine can be designed by combining the standardized modules. Here, a module is a simple hardware or software unit that performs a simple function and can be easily combined with others. This paper proposes two design principles of such machines from a viewpoint of information. The first design principle is the consideration of the system's entropy. Entropy can be a measure of the uncertainty of the system and used as a criterion for evaluating module combination. The second is the consideration of information flow balance in the system.Interpretive Structural Modeling (ISM) is useful to clarify the inherent hierarchical structure in the information system. After ISM, the information processing load and data transfer rate in the system can be examined. The combination that decreases the entropy with a minimum information processing load is preferable. Experiments to design a sensor system were conducted and a combination of a CCD camera and an ultrasonic sensor demonstrated the usefulness and effectiveness of the principles.