This paper proposes a robotic eye movement control system based on neural pathways of human binocular motor system. The ability so called “Translational Vestibulo-Ocular Reflex (TVOR)” in physiology area of eye movements is developed to compensate the motion bular caused by the translation head movement. Since TVOR depends on both the movements of the head and the position of the related target with the head, it cannot be realized as a single eye system. According from the physiology and anatomy studies, its target position information is from the angles of both eyes in ocular motor system of human. In this paper, the Smooth Pursuit eye movement, Rotational Vestibulo-Ocular Reflex (RVOR), and TVOR are successfully combined in one binocular motion control model. This model has been approved through the experiment by using a mobile robot with binocular active cameras, accelerometer, and gyroscope. Its result shows that the cameras can keep tracking on a moving target smoothly while the robot is violently moving. This ability cannot be realized by only using visual feedback control method.
Robots that interact with humans in household environments are required to achieve multiple simultaneous tasks such as carrying objects, collision avoidance and conversation with human, in real time. This paper presents a design framework of multiple human-interacting tasks to meet the requirement by considering stochastic behavior of humans. The proposed designing method first introduces petri-net for parallel multiple tasks. The petri-net formulation is converted to Markov decision processes and processed in optimal control framework. Multiple task arbitration is resolved by optimization with approximated value functions. Two tasks of safety confirmation and conversation tasks are mutually interacting and expressed by petri-net. Tasks that normally tend to be designed by integrating many if-then rules can be dealt with in a systematic manner in the proposed framework, that is, in a state estimation and optimization framework. The proposed arbitration method was verified by simulations and experiments using RI-MAN, which was developed to do interactive tasks with humans.
In this paper, we propose a new robotic catapult with a high stiffness endpoint. The conventional robotic catapults based on the closed elastica are the robotic elements for generating impulsive motions by utilizing the snap-through buckling. In a typical closed elastic catapult, the two ends of an elastic strip are fixed to a free joint and an active joint, respectively. Here we found that by adding only the high stiffness at the free joint, compared to the conventional type, more elastic energy can be stored and release surely without loss of a characteristic of generating impulsive motion repeatedly. By utilizing the proposed robotic catapult based on the closed elastica with a high stiffness endpoint, we develop a novel swimming robot which can swim not only continuously but briskly from a stationary state.
A conceptional idea of the pseudo mechanism is proposed. This concept means that a different mechanical feature from the natural one appears on the mechanism by means of control method. For example, a worm gear behaves like an one-way clutch. Generally, worm gears cannot be driven by load torque because of its mechanical stop feature. But the worm gear of this concept can be driven by load in specified direction, and the rotation to the other direction is prohibited. This asymmetrical motion is achieved by a switching operation of a force feedback control. The applied load torque is detected and fed back positively to a position servo control system in order to achieve power assist control. The worm gear seems to be driven by the load torque and it resembles to a spring motion. If the force feedback line is cut off, the drive system cannot be driven by a load. The switching operation of the force feedback line causes to a selection of back drivable feature or mechanical stop feature. The switching condition is set by a direction and threshold value according to the required mechanical feature. The pseudo mechanical worm gear also pretends as a torque limiter or a ratchet.
This paper proposes a method to determine a contact point location by a 6D force sensor with noise on force/moment signals. In the noiseless case, it is well known that candidates of the contact point are represented by a line parallel to the acting force, and the contact point can be determined from the intersection of the line and the robot surface. However, if the line is constructed from actual force/moment sensor signals directly, the candidates of the contact point may be unreliable because of the measurement noise. In this paper, we propose a method to determine the contact point from an optimization problem, in which the cost function represents the error of the force/moment balances. Candidates of the contact point are also represented by a line, which is parallel to the estimated acting force in this case. As in the noiseless case, the contact point can be determined from the intersection of the line and the robot surface. Numerical examples and experimental results are shown to prove efficiency of the proposed method.
The purpose of our paper is to design a three dimensional passive walker with ankle springs and flat feet. Some 3D passive walkers have been developed with circular-arc or spherical foot with some passive devices to compensate yaw and roll motions. However, the foot might slip about yaw axis because of low friction. We propose a flat foot with ankle springs which stabilizes the roll motion, and give a design method for the spring stiffness. Moreover, we prospect an effect of synchronization between the swing leg and lateral motion. Finally, we verify its effectiveness through experiments.
In this paper, we propose a fast and accurate scene recognition system from moving vehicle, which is a combination of Higher-order Local Auto-Correlation (HLAC) features and the linear discriminant analysis. The scene recognition system of moving vehicles should classify successive images into categories under limited computational resources, have robustness to cope with environmental changes, and adopt various scenes, which do not have explicit features. Our algorithm satisfies those requirements. We evaluated our algorithm in intersection and highway recognition experiments, and showed that our algorithm not only has comparable recognition accuracy compared with the state-of-the-art methods but also outperforms those methods in computational speed.