抄録
This paper is intended to intensively discuss how intelligence of real-time adaptability can be designed from perspective of "embodied system". One of the significant features of our approach is that we explicitly exploit an "emergent phenomenon" stemming from the interaction between control and mechanical dynamics in order to control locomotion in real time. To start from simplicity, we focus attention on locomotion of a protozoan organism, e.g., slime mold and amoeba, which exhibits remarkably intelligent behaviors, such as avoiding hazardous condition, approaching nutrients and humidity, and solving a maze. These behaviors are based on amoeboid movement, which is driven by the flexible epitheca and the protoplasm in the absence of a central nervous system or specialized organs. In light of these facts, we made a simulation model, particularly focusing on epitheca consisting of "real-time tunable springs" and "low of conservation of protoplasmic mass", the former of which is used to deal with local interaction dynamics and the latter of which is used to deal with global interaction dynamics. Simulation results indicate that the proposed model can induce highly adaptive locomotion according to the situation without losing the coherence of the entire system.
