- Collective motion
Considering the physical properties of active elements with mutual interactions, we proposed a mathematical deterministic model of collective motion based on Newtonfs equation of motion for particles and the heading dynamics. Several kinds of cluster motion such as marching, wandering, and swarming emerges by the model. <detail>

Research Interest

My research interest is to realize a collective intelligent system, which consists of interacting simple agents and shows advanced functions such as complex motion, cooperative transportation and structure formation.

- Task allocation for multi-robot
We proposed a dynamical model for multi-states task allocation. It consists of identical elements and external variable which is inspired by a polyethism of social insect. We mainly discuss it from the viewpoint of proportion regulation of a population between different states. We demonstrate its performance by numerical simulation and robot experiment.<detail>
- Formation structure
We proposed an algorithm of self-organizing formation by using virtual springs among the elements. The element generates virtual springs between neighbor element based on information how many other elements exist in neighborhood with a certain radius. Although the elements interact locally only by virtual springs, they form a desired shape much larger than the sensory radius.<detail>

- Ants
Ant is a highly suggestive creature for developing collective robotic system. In order to obtain some clues for advanced robotic system, we are measuring the behavior of ants. We found single ant has a certain rhythm in its active-inactive state. Now we consider the single ant as a motile oscillator are studying their collective characteristic from physical viewpoint of coupled oscillators.<detail>