This study focuses on the derivation of elastic force models of a hemispherical soft fingertip and reveals the existence of a local minimum of elastic potential energy (LMEE) induced by the deformation of the fingertip. The proposed model can be written in a straightforward equation when the geometric nonlinearity of the fingertip is applied in the derivation process, which is suitable for analyzing the stability of robotic handling system with soft fingertips. We also show complete elastic force and potential energy equations by incorporating the material nonlinearity into the derived model. Furthermore, we compare our models with the Hertzian contact theory that is applicable to the practical usage for the strength design of a lot of machines. Finally, we demonstrate the validity of our models by comparing simulation results and experimental results.

View full abstract

This paper proposes a new method of locomotion control named “Stiffness Distribution Control (SDC) ” to realize various locomotion of a closed loop robot with mechanical softness on the ground. SDC directly defines a reference stiffness coefficient at each hinge by a SDC string, which represents distribution of stiffness coefficient. This simple method does not need costly calculation at all and thus is suitable for controlling closed loop robots. We build a closed loop robot “BIYOn” with newly designed variable stiffness hinges. SDC is experimented on the robot and effectiveness of SDC is verified in practice. A two-stage systematic optimization approach for SDC strings, which is composed of “hinge-wise optimization method” and “collective descent optimization method, ” is also developed to obtain efficient locomotions. The former is the method focused on functions of each hinge in desired locomotion, and the latter is to optimize the solutions based on local solutions ontained in the first stage. This approach does not require a large amount of computation to get global optimal SDC strings. By this approach, we succeed in obtaining a fast and smooth rolling motion and a sudden stopping motion.

View full abstract

This paper proposes a system that practically identifies the position and orientation of hand bones from magnetic resonance (MR) volume images by registration of a bone model. Investigation of the link structure of a human hand requires the measurement of the relative movement of the bones by identifying each bone configuration among a large number of hand poses of the same subject. The proposed system aims to save total time in acquiring bone configuration with the same accuracy compared with the completely-manual segmentation. From one of the MR scanning data of different poses. a bone model is generated by manually segmenting the bone region. The initial configuration of the bone model is set interactively by an operator and the model is aligned in MR voxel data by minimizing the performance index. Experimental results show the validity oft he proposed system in terms of accuracy and required time for data-processing.

View full abstract

We describe a memory system of motion patterns which forms symbolic representations and maintains specificsymbolic hierarchy based on stored memories. The system consists of auto-correlation based feature vectors and a nonmonotonic associative memory model. The feature vectors clarify the global structure of motion patterns as a cluster structure. The associative memory model encodes symbols and the hierarchy by bifurcations attractors naturally mirroring the inherent structure of the stored motion patterns.

View full abstract

The traveling operation of a mobile manipulator involves some difficulties to control, such as dynamical interference and non-holonomic character. The integration of errors in trajectory-tracking of the mobile manipulator is caused by the non-holonomic constraint expressed by a differential equation of first order. The integrated error never converges to zero by a controller being used usually for fixed manipulator. Constructing a controller including a consideration of the non-holonomic constraint can solve this problem. We propose a control method that guarantees zero convergence of guidance errors and the trajectory tracking errors of the mobile manipulator. The proof is given by Liapunov method. Simulations confirm the performances of the proposed controller, and the results show that the guidance errors and trajectory-tracking errors converge to zero. Furthermore how the dynamical influences caused by slipping of carrying objects on the mobile manipulator affect the trajectory-tracking performances of the proposed controller is evaluated by several simulations.

View full abstract

Remote diagnosis can be realized using communication network. We have been developed a master-slave type remote medical system for the diagnosis of the shoulder disease such as dialysis related amyloid arthropathy (DRAA) by ultrasonographic images. Proper position, orientation, and contact force between the ultrasound probe and the affected part of the patient should be required to acquire proper diagnostic images in the diagnosis. Safety and manipulability are also required to construct the remote medical system through the communication network. Then, it has impedance control capability for the master and slave manipulators' positions to display contact force and enhance manipulability. And it has Continuous Path control capability for the slave manipulator's orientation to realize the smooth and accurate motion of the ultrasound probe even if it is difficult to transmit control data at high sampling rate through the communication network. The results of the remote diagnostic experiments demonstrated that the healthcare professionals could perform the diagnosis for the real patients through the communication network using the constructed system.

View full abstract

This paper proposes a method for constructing a simulator of pinching a 3-D object by a pair of robot fingers under the gravity effect, holonomic constraints, and a differential equation of non-holonomic constraints expressing rotational motion of the pinched object. A noteworthy difference of modeling of motion of a 3-D object from that of a 2-D object is that the instantaneous axis of rotation of the object is fixed in the 2-D case but that is time-varying in the 3-D case. This difference appears as non-holonomic constraints. A further difficulty 3-D physical interactions between two robot fingers and a rigid object is that spinning motion may arise around the opposing axis connecting the two contact points between fingertip's and the pinched object. The proposed simulation method deals with the case after such spinning motion ceases. Non-occurrence of such spinning motion induces a further non-holonomic constraint. It is shown that Lagrange's equation of motion of the overall system can be derived without violating the causality that governs such non-holonomic constraints. In the mathematical model, holonomic constraints and Euler-Lagrange equations are mutually related. Therefore, this paper aims at ordering these conditions and proposing the algorithm to carry out numerical simulation of the overall system motion without violating the principle of causality of time. Then, results of the simulation are shown to verify the validity of the proposed control input called“blind grasping”based on opposable forces between the thumb and another finger (index or middle finger) without use of any object kinematics or any external sensing.

View full abstract

This paper presents a motion suspension system to suspend humanoid motion in case of emergency. Once humanoids start their motions in human daily environments, there is a possibility that humanoids will meet with several emergencies such as hurting humans and injuring themselves. Even so. humanoids should be controlled so that they avert such emergencies in real-time. To realize this demand, we propose a method of real-time judgment of emergency prediction by humanoids. We also propose a simple and effective method of real-time pattern generation to force humanoids to stop immediately by one step without falling. To verify the validity of the proposed method, we finally present experimental results using a humanoid robot HRP-2, which include experiments at 2.8 [km/h] walks.

View full abstract

This paper proposes a linear motion mechanism for a robot gripper that can exert a large grasp force. A robot gripper requires speed when opening and closing its jaws, and force when grasping an object. The proposed linear motion mechanism can drive its output quickly to an object and can generate a large grasp force by using a lever. The mechanism has two feed screws with different pitches along one screw shaft. Initially a gear is fastened on the feed screw with the smaller pitch to rotate the shaft. The feed screw with the larger pitch enables quick motion of the output nut. After the output makes contact with an object, the rotation of the shaft stops. By further driving the gear, it becomes loosened and begins to move axially by the feed screw. The gear pushes the lever, which generates a large fingertip force. We have developed an 86g gripper with this mechanism and experimentally verified that it can grasp objects of various sizes and can exert a large grasp force over 200N. It is also experimentally verified that the gripper can grasp and ungrasp an object reliably over 5000 times if machine lubricant is properly supplied.

View full abstract

This paper is concerned about a variable impedance control for human-machine cooperative positioning systems. The prospective profile is found from the dynamic equation and the properties of the operational profiles known as suitable ones. We propose a simple variable impedance control in which the damping coefficient is varied according to the operational force of human based on the above profile. The experiments of positioning tasks are examined and the effectiveness of the proposed method is shown. The properties on conventional methods are compared.

View full abstract

Geometric haptic rendering algorithms, in which virtual proxy positions are determined according to input device positions, are widely used for impedance-type haptic rendering. Such methods are intuitively simple because they reduce mechanistic problems of determining reaction forces into geometric problems of updating proxy positions. Aiming to extend the scope of application of geometric haptic algorithms, this paper presents a mechanistic interpretation of geometric algorithms and proposes two techniques. One is a technique to combine geometric algorithms and viscous damping in impedance-type haptic rendering. The other is a technique to utilize geometric algorithms in admittance-type haptic rendering. The results of implementation experiments are presented.

View full abstract