Masanao Koeda

Shape recognition and grasping by robotic hands with soft fingers and omnidirectional camera

The purpose of this paper is to establish a method of shape recognition and grasping of unknown objects using a robotic hand with two soft fingers and one omnidirectional camera. For shape recognition, we propose to use a simple 2D version of the visual volume intersection method which takes advantage of a special feature of the omnidiretional camera attached to the hand. For grasping we propose a simple grasp quality criterion to obtain the best grasping position for the two soft fingers based on a visual hull in a horizontal grasp plane. We conducted several experiments and the results show the validity of the proposed method.

Shuffle Turn with Both Feet of Humanoid Robot by Controlling Load Distribution of Soles

This paper proposes a novel shuffle turning method for a humanoid robot by controlling load distribution of each sole. Conventionally, turning motion of a humanoid robot is performed by repeating of stepping the feet. However the motion is inefficient and time consuming. In our proposed methods, the feet are slipped on the floor without stepping. To minimize variation of the turning angle due to the friction variation of the floor, the directions of the feet are changed following a predetermined trajectory and the load distributions of the soles are controlled to be nonuniform. Experiments with the humanoid robot HOAP-2 were conducted to verify the validity of the proposed method.

Shape Recognition and Optimal Grasping of Unknown Objects by Soft-Fingered Robotic Hands with Camera

In this paper a method is proposed for shape recognition and grasping of unknown objects using a robotic hand with two soft fingers and one omnidirectional camera. For shape recognition, we propose to use a simple 2D version of the visual volume intersection method to obtain the visual hull of the object in a horizontal plane from multiple images taken from different positions around the object. For grasping we propose a simple grasp quality criterion to obtain the best grasping position for the two soft fingers based on the obtained visual hull. We conducted several experiments to show the validity of the proposed method.

Protosymbol emergence based on embodiment: robot experiments

Robotics can serve as a testbed for cognitive theories. One behavioral criterion for comparing theories is the extent to which their implementations can learn to exploit new environmental opportunities. Furthermore, a robotics testbed forces researcher to confront fundamental issues concerning how internal representations are grounded in activity. In our approach, a mobile robot takes the role of a creature that must survive in an unknown environment. The robot has no a priori knowledge about what constitutes a suitable goal-what is edible, inedible, or dangerous-or even its shape or how its body works. Nevertheless, the robot learns how to survive. The robot does this by tracking segmented regions of its camera image while moving. The robot projects these regions into a canonical wavelet domain that highlights color and intensity changes at various scales. This reveals sensory invariance that is readily extracted with Bayesian statistics. The robot simultaneously learns an adaptable sensorimotor mapping by recording how motor signals transform the locations of regions on its camera image. The robot learn about its own physical extension when it touches an object. But it also undergoes an internal state change analogous to the thirst quenching or nausea producing effects of intake in animals. This allows the robot to learn what an object affords by relating these effects to learned clusters of invariance. In this way primitive symbols emerge. These protosymbols provide the robot with goals that it can achieve by using its sensorimotor mapping to navigate, for example, toward food and away from danger.

A Quantitative Evaluation Method of Handedness Using Haptic Virtual Reality Technology

This paper proposes a method for evaluating quantitatively the degree of handedness of a person by using the haptic virtual reality technology. A definition is also given of a pair of quantitative indices for handedness and dexterity of a person from the data obtained by this method using the factor analysis. A haptic display system has been developed consisting of two PHANToM OMNI haptic devices, a graphics display, and a computer for virtual space calculation and control. To observe the skillfulness of two hands of a person, three virtual tasks have been developed: positioning task, force control task, and manipulation task. Ten male subjects aged from 22 to 23 years old have taken the test. The obtained performance data from these tasks have been analyzed using the factor analysis. Since the obtained factor scores for the right and left hands of each subject can be regarded as the skillfulness of the right and left hand, it is proposed to define the degree of handedness of the subject based on the difference of these factor scores. The proposed degree of handedness and the result of conventional LQ method did not contradict to each other. The calculated degrees of handedness, however, differ largely among the subjects. From this result, it is expected that the proposed method can be useful for detailed and quantitative evaluation of handedness.

Stable Grasp and Manipulation in 3D Space with 2-Soft-Fingered Robot Hand

We report on our newly developed two-fingered robotic hand with soft and thin skin. This hand is basically controlled by a position-based control algorithm. However, it can realize desired grasping force due to the soft skin while maintaining stability of the system. For an experimental validation of this concept, we conducted following the three kinds of manipulation tasks: manipulating of a cubic object, opening and closing of a bottle cap, and turning of a crank. These results show that this two-fingered hand can perform various manipulation tasks of objects in three dimensional space.

2A1-D20 Two Footed Shuffle Turn of Humanoid Robot by Controlling Load Distribution of Soles

This paper proposes a novel turning method for a humanoid robot by controlling load distribution of each soles. Conventionally, turning motion of a humanoid robot is performed by repeating of stepping the feet. However the motion is inefficient and low stability. In our proposed method, the feet are shuffled on the floor without stepping. To minimize the effect of the friction of the floor, the load distributions of the soles are taken to be nonuniform by controlling the angles of the feet. To verify the feasibility, experiments with the humanoid robot HOAP-2 were conducted and the results showed that the nonuniform load distribution is effective for correct turning.

2P1-F08 Humanoid Home Robot System with Speech and Image Recognition

This paper describes a humanoid home robot system with speech and image recognition.Humanoid home robots are expected to work in human living environment,and are needed to have functions for easy operation.To perform object searching, approaching and carrying task,we implemented functions of speech recognition, object recognition by single camera, and speech response. A preliminary experiment was conducted and its feasibility