Sho Yamauchi

Conquest Oriented Robot Knowing Its Own Availability

Robot is designed for achieving specific task. However, robot is able to do much more things than it is expected by its own hardware. In this paper, concept of conquest oriented robot is proposed to know its availability systematically and autonomously. Algorithm of conquest oriented robot and its experimental result in simulation field for the first step is shown in this paper.

Interaction of Two Oscillator Aggregations

The main feature that keeps states and structures stable can be seen in living organisms. This adjusting and adaptive features are called homeostasis. This integrated adaptive feature is achieved by the cooperation of organs in living organisms. Living organisms in nature act dynamically due to this feature. Highly adaptive behavior caused by this feature is also observed in simple living organisms that have no neural circuits such as amoebas. Based on these facts, a method of control to generate homeostasis in robotic systems is proposed by assuming a robot system is an aggregation of oscillators and each parameter in a robot system is allocated to an oscillator. Especially, interaction between two independent robots as oscillator aggregation is focused in this paper.

Oscillator aggregation wheeled robot actively synchronizing with external environment

The main feature that keeps states and structures stable can be seen in living organisms. This adjusting and adaptive feature is called homeostasis. This integrated adaptive feature is achieved by the cooperation of organs in living organisms. Living organisms in nature act dynamically due to this feature. Highly adaptive behavior caused by this feature is also observed in simple living organisms that have no neural circuits such as amoebas. In amoeba case, each cell acts as an oscillator and such behavior is explicable as synchronization phenomena of oscillators. Based on these facts, a method of control to generate homeostasis in robotic systems is proposed by assuming a robot system is an aggregation of oscillators in this paper. When a robot system is assumed as an aggregation of oscillators, a robot system tries to keep the value of the specific function that indicate its uncomfortable level small the whole time. To keep such function value small, a robot system stabilize the relationship between oscillators when its value is small and reconstruct the relationship between oscillators when its value is large. This behavior is also explicable as synchronization phenomena of oscillators. Wheeled robot with camera is applied this control method in this paper to confirm the effect of the method. Adaptive behavior is observed in the robot through the task to find the better position between the object.

Arbitrary Lattice Formation with Flocking Algorithm Applied to Camera Tracking System

Flocking algorithms for a multi-agent system are distributed algorithms that only have simple rules for each agent but generate complex formational movement. These algorithms are known as swarm intelligence and are robust and disaster tolerant for most cases.We consider that flocking algorithms that have these characteristics are the way to generate homeostasis in a system.We expect that by making use of this algorithm the system can tune its self parameters and thus maintain a high performance. First, to apply a flocking algorithm to a system, we extended the flocking algorithm to form an arbitrary lattice for further flexibility.We then applied the extended flocking algorithm to a position tracking camera system as an example.

Oscillator Aggregation in Redundant Robotic Systems for Emergence of Homeostasis

The main feature that keeps states and structures stable can be seen in living organisms. This adjusting and adaptive feature is called homeostasis. This integrated adaptive feature is achieved by the cooperation of organs in living organisms. Living organisms in nature act dynamically due to this feature. Highly adaptive behavior caused by this feature is also observed in simple living organisms that have no neural circuits such as amoebas. Based on these facts, a method of control to generate homeostasis in robotic systems is proposed by assuming a robot system is an aggregation of oscillators in this paper and each parameter in a robot system is allocated to an oscillator. Such oscillators interact so that the whole system can adapt to the environment. Also, a redundant robot arm is made to confirm the effect of this control method to generate homeostatic behaviors in robotic systems.

Application of Flocking Algorithm to Attitude Control of Humanoid Robot

Flocking algorithm of multi agent system is robust and disaster tolerant even though agents face obstacles or destructed. In this paper, we focus on applying these mechanism to single humanoid robot by assuming that the robot is the set of agents. We extend flocking algorithm to be capable of using in single robot and apply it to humanoid robot to maintain standing posture.

Extended formation algorithm for combination of subgroups

Flocking algorithms for a multi-agent system are distributed algorithms that only have simple rules for each agent but generate complex formational movement. These algorithms are known as swarm intelligence and are robust and disaster tolerant for most cases. On the other hand, most of organizations are constructed from smaller groups, we call it subgroup here. It means activities of organization should be determined by its subgroups but flocking algorithm itself does not have such feature. To express this feature, we extend flocking algorithm to be able to contain subgroups of agents. Also we add flocking algorithm the term that indicates the relationship with other subgroups. By this extension, flocking algorithm is be able to form specific shape more flexibly.