Generally, large knee joint torque is required when a leg is flexed. However, the large torque motor will increase the robot size and total weight. Also, a large reduction ratio gear box to realize the large torque will decrease output speed and lower backdrivability of the joint. It makes the robot difficult to perform agile and flexible motion like animals. In this paper, we propose a robot leg with a knee joint mechanism consisting of a variable reduction ratio crossed four-bar linkage mechanism (VRRCFLM) based on cruciate ligament of an animal. The VRRCFLM is to increase the reduction ratio for large knee flexion postures without greatly reducing the total backdrivability. In this paper, we developed a robot leg with a knee joint mechanism consisting of the VRRCFLM. In order to design the link parameters of the leg mechanism, optimization design aimed at maximizing the jumping height of the robot was performed. The robot model with the designed mechanism was evaluated through the dynamics simulations. Thanks to the VRRCFLM, the required torque of the knee motor at the large flexion postures was decreased. Moreover, the vertical jumping height was improved by 24.6 % comparing with a model without the mechanism. In experiments using the prototype, the required static torque was decreased as in simulation, and the jumping height was more than one leg length.

The robustness of the positioning and posturing of robot hands relative to target object and support surface is an important issue for autonomous grasping. For example, to perform a grasping action such as picking up thin objects from a table top, the position and posture of the hand must be controlled to keep adequate relative posture and distance to the support surface besides those between the hand and the target object. Because slight errors in the posture and position are enough to cause grasping failure, the positioning and posturing of the hand must be precise enough, specially when the hand is close to the target object and support surface. To improve the robustness of robotic grasping, in this paper we present a method by grasping control based on the relative posture and position between hand and support surface besides those between hand and target object, using proximity sensors. Proximity sensors are newly installed on fingernails besides on the fingertips. As the fingernail sensor, an integration of Time-of-Flight (TOF) sensor and photo-reflector is designed to realize long range detection, as well as with precise and high-speed detection regardless of the reflectance of support surfaces when approaching the support surface. By the sensors, the hand can approach the object and support surface coarsely first, and then can be controlled fast and precisely to realize adequate grasping motion along the support surface but without contact with the support face. The method has been implemented to a manipulator system, and successful grasping experiments have demonstrated the effectiveness of the proposed method.

In this study, we propose integrated control of a robotic hand and arm using only proximity sensing from the fingertips. An integrated control scheme for the fingers and for the arm enables quick control of the position and posture of the arm by placing the fingertips adjacent to the surface of an object to be grasped. The arm control scheme enables adjustments based on errors in hand position and posture that would be impossible to achieve by finger motions alone, thus allowing the fingers to grasp an object in a laterally symmetric grasp. This can prevent grasp failures such as a finger pushing the object out of the hand or knocking the object over. Proposed control of the arm and hand allowed correction of position errors on the order of several centimeters. For example, an object on a workbench that is in an uncertain positional relation with the robot, with an inexpensive optical sensor such as a Kinect, which only provides coarse image data, would be sufficient for grasping an object.

Cats have protractible claws to fold their tips to keep them sharp. They protract claws while hunting and pawing on slippery surfaces. Protracted claws by tendons and muscles of toes can help cats anchoring themselves steady while their locomotion trends to slip and releasing the hold while they retract claws intentionally. This research proposes a kind of modularized self-adaptive toe mechanism inspired by cat claws to improve the extremities' contact performance for legged robot. The mechanism is constructed with four-bar linkage actuated by contact reaction force and retracted by applied spring tension. A feasible mechanical design based on several essential parameters is introduced and an integrated Sole-Toe prototype is built for experimental evaluation. Mechanical self-adaption and actual contact performance on specific surface have been evaluated respectively on a biped walking platform and a bench-top mechanical testing.

This study presents a robot-hand-arm system with high robustness and responsiveness by using a “Net-Structure Proximity Sensor.” The sensor, which we have developed and specially designed for a robot hand, directly detects an object being to be grasped and outputs analog voltage signals according to the position/posture error between the robot hand and the object. It has been confirmed that the robot hand is able to quickly adjust to and grasp an unknown object by applying a feed-back control method based on the sensor signals. This paper focuses on the integration of the proximity-based feedback control to a commonly-used vision-based control. These sensors work in complementary manner: a vision sensor is available for planning an approaching path of a robot hand by detecting large area, and a Net-Structure Proximity Sensor enables the robot hand to adjust the approaching error before grasping and to improve the certainty of the grasping. Two objective velocities are derived independently by the sensors. By adding the velocities with considering the reliability of the sensor information, the robot hand becomes to be able to perform approaching and adjustment to the target object simultaneously. Experimental results showed that the robot hand grasped a moving object with high success rate even in conditions where it was difficult to predict the trajectory of the object accurately.

Quick motion and soft touch control are important for autonomous grasping. To perform a grasping action, a hand must adjust its fingertips to match the object shape. It is also necessary to reduce the fingertip velocity on contact with the object. We propose a method of fingertip velocity control for fast approach and slow contact using proximity sensors installed on fingertips. The proposed control reduces the fingertip velocity at contact, decreasing the change of impulse force and thus making it appropriate for grasping soft or fragile objects. The proposed control uses time-to-contact (TTC), which is converted from sensor output. It is known that many animals, including humans, use TTC for collision avoidance. TTC represents the remaining time until collision considering the rate change of a control variable. We carried out grasping tests on various common objects using TTC. Experimental results show that the proposed control realizes fingertip alignment perpendicular to unknown shapes and surface objects and lowers the velocity at contact.

Within the limbs of typical animals, there exist bi-articular muscles crossing two joints. It is known that the bi-articular muscles of the felid play an important role in the locomotion. Also the muscle-tendon complex, composed of the gastrocnemius muscle and the Achilles' tendon that cross the knee joint and the ankle joint contributes much to the movements such as running and jumping particularly. Besides, because the muscle-tendon complex has the function for absorbing shocking, it is utilized for soft landing from high places. To achieve high performance for jumping and landing motion like cats, we are developing robot legs inspired by the bi-articular muscle-tendon complex of cats. The leg consists of hip, knee and ankle joints. For the knee and ankle joints, a four-bar linkage mechanism with one elastic linkage, in which the knee joint is driven by an electric rotary motor and the ankle joint is passive, is applied. By this mechanism, basic functions of the bi-articular muscle-tendon complex of felids like cats can be realized and the performance for jumping and landing can be improved. In this paper, the new leg mechanism is described. Moreover, a prototype of a pair of the hind legs of the quadruped robot using the new mechanism has been developed. The results of jumping and landing experiments are shown to validate the effectiveness of the mechanism.

Aiming to design a soft robotic fish with more natural, more flexible and high-performance movements through biomimetic method, we are developing a soft robotic fish with body/caudal fin (BCF) propulsion by using piezoelectric fiber composite (PFC) as actuator. Compared with conventional rigid robotic fish, the design and control of the soft robotic fish are difficult and hard to reveal its dynamic performances due to the large deformation of flexible structure and complicated coupling dynamics with fluid. That's why the design and control method of the soft robotic fish has not been established, and we need to study it by considering the interaction between flexible structure and fluid. In this paper, fluid-structural coupling analysis based on acoustics method is applied to consider the fluid effect and predict the dynamic responses of soft robotic fish using PFC in fluid. Basic driving and governing equations of soft robotic fish in the fluid are firstly described. Then the numerical acoustics coupling analysis is performed. The calculated results are congruent well with experiments on dynamic responses. Finally, based on this coupling method, a new prototype of soft robotic fish is proposed by optimization for improvement.

In this paper, we demonstrate a preliminary motion before grasping by a robot hand, for adjusting the object-fingertip distance and 2-axis postures simultaneously, using a Resistor Network Structure Proximity sensor (RNSP sensor). Through this motion (called “pre-shaping”) and the grasping of an object, the surface of each fingertip is brought into contact with the object surface so that in the next stage grasping can be undertaken. In the next stage, a force can be applied from the fingertips onto the object surface directly. The pre-shaping enhances the reliability of the feedback control for the after-contact tactile sensors. To realize the pre-shaping, we use fingertips equipped with RNSP sensors, which can detect the distance between the fingertip and the object, to determine the relative position between fingertips and an object. The RNSP sensor has a fast response (<;1 [ms]) and simple connectivity (only 6 wires), and can be mounted easily. Additionally, a characteristics of the RNSP sensor output can be designed by the arrangement of the sensor elements. To perform the pre-shaping by simple sensor feedback control based on the configuration between the fingertip and object, we designed the RNSP sensor so that it had the appropriate characteristics for the pre-shaping.

It is well known a fine tuned unpowered biped machine can passively walk down a gentle slope. The in-depth understanding of the passivity is helpful to design more efficient bipedal gaits. In this paper, we investigate the principle mechanism of passive dynamic walking and propose a unified framework for virtual passive bipedal gait generation. The framework shows the physical system's kinetic energy varies inversely proportional to the imitated virtual system's potential energy. Based on this property we propose a kinetic energy tracking control law to realize virtual passive dynamic walking on any slopes in any virtual gravity fields. Numerical simulations are given to support our proposal.

This paper describes a combination mechanism of a wheeground a paddle to realize an effective movement on the border of ground and water. For the conventional mobile mechanisms on water or on ground the combination of the mechanism to the body of the robot separately, and each one can be the interruption to the other in each field. In this research, we propose a hybrid mechanism of a wheeground a paddle as drive mechanism itself to realize an effective movement on the border such as a coast with debris after a disaster like a seismic sea wave or a tanker grounding accident. A prototype robot has been built and basic experiments on its motion have been conducted.

Tactile sensors for slip detection are essential for implementing human-like gripping in a robot hand. In previous studies, we proposed flexible, thin and lightweight slip detection sensors utilizing the characteristics of pressure-sensitive conductive rubber. This was achieved by using the high-frequency vibration component generated in the process of slipping of the gripped object in order to distinguish between slipping of the object and changes in the normal force. In this paper, we design a slip detection sensor for a multi-fingered robot hand and examine the influence of noise caused by the operation of such a hand. Finally, we describe an experiment focusing on the adjustment of the gripping force of a multi-fingered robot hand equipped with the developed sensors.

This paper aims at mobility enhancement of flexible underwater robots. For this purpose, a novel propulsion method utilizing both static and dynamic deformation of a flexible planar structure is proposed. By actuating the planar structure in two dimensions, various propulsive motions can be generated and the realization of various robotic movements can be expected. Two prototypes of such kind of robots with square and circular shape driven by two pieces of piezoelectric fiber composites has been developed. As the experimental results, the movements with multiple DOFs, holonomic and non-holonomic movements have been realized by the robots. The motion mechanism for the movements are discussed by simulation results.

This paper describes linear load-sensitive continuously variable transmission with the spherical driving unit. This CVT mechanism consists of spherical drive, drive axis, motor housing, fixed bracket and linear sliding plate. It changes the reduction ratio continuously by inclination angle of active rotational axis. Additionally, this linear mechanism has a load-sensitive function by changing inclination of active rotational axis in response to the load. We have developed a linear load-sensitive continuously variable transmission and confirmed the effectiveness of the proposed mechanism.

This paper presents a proximity sensor for recognizing an object in 360 degrees all around with response time only 1[ms]. The sensor applies an unique design of analog net-structure circuit, which enables all-face mounting on various peripheries and fast-response/simplification of information processing. Thus, the sensor is especially available for avoiding sudden contact with rapidly approaching objects for mobile robots, robot arms, robot hands, and other robotic systems. In this paper, the structure and the detecting principle of the sensor are firstly described, then, its detection property is tested by experiments. The result shows that the sensor recognizes the angle of the nearby object with accuracy of 3[deg]. Furthermore, as practical use of the sensor, a large size model is produced for a mobile robot to achieve omni-directional sensing. By using the application model, two detection methods are discussed on situations where the number or the positions of the nearby objects differs.

To achieve the skillful task like the human, many researchers have been working on robot hand. An interaction with vision and tactile information are indispensable for realization of skillful tasks. In the existing research, the method using a camera to get the vision information is often found. But, in the boundary area of a non-contact phase and a contact phase, there are problem that lack of sensor information because the influence of occlusion comes up to surface. We devise to introduce the proximity sensor in this area. And we call the robot hand which is equipped with proximity, tactile and slip sensor “intelligent robot hand”. In this research, we show the constitution example of the intelligent robot hand and propose the method to realize Pick&Place as concrete task.

Slip-detecting tactile sensors are essential for achieving human-like gripping motion with a robot hand. In previous research, we developed a flexible, thin and lightweight slip sensor that exploits the characteristics of pressure conductive rubber. However, using this sensor, it was difficult to distinguish between object slip and a change in the normal force. Therefore, in the present research, we investigated a method for identifying object slip by analyzing the frequency components of the output signal from the sensor. As a result, we found that high-frequency components of 1 kHz or more are included in the complex voltage signal generated by object slip. Therefore, by using this high-frequency component, we developed a simple sensor that distinguished between both contact and initial slip with high sensitivity.

Recently, flapping robots are noticed by many researchers, and the realization of flight motion like insects, birds and so on has many challenges. The purpose of this work is to develop flapping robots using a new type of piezoelectric material, that is, piezoelectric fiber composites. By using the composites, actuating as well as energy harvesting and sensing fibers can be embedded into the wing as part of the structure, which makes the robot flap while remaining compact and light. This paper describes the performance enhancement of the flapping robots capable of flapping and feathering motions using piezoelectric fiber composites. A unique mechanism with different stiffness for up-stroke and down-stroke and a new drive control using saturated sine wave are proposed in this paper. By introducing the mechanism and drive control, the performance about lift and thrust force has been improved remarkably.

Fish type underwater robots has a lot of possibilities such as good mobility and high efficiency. Most of the fish type robots developed up to now have complicated mechanisms with complicated motion control. It results in that the structure and the movement of robots are different from that of fishes. The purpose of our work is to develop creature-like flexible underwater robots by using piezoelectric fiber composite. Compared with conventional artificial muscular underwater robots, powerful underwater robots with a very simple structure can be composed. This paper describes the development of a flexible underwater robot with caudal fin propulsion, which can generate large propulsive force and move at high speed. Besides, by utilizing the large propulsive force, pitch and roll motions are realized by attaching two pectoral fins to the robot.

This paper describes various configurations of two connected unit crawlers. By changing the relative position of the two connected vehicle units, the overall robot comprising such a mechanism can automatically adapt to surface obstacles on the field, including complicated structures such as disaster-generated debris. In addition, we analyzed the effect of axis arrangements in order to simplify the realization of the switching function so as to achieve the four basic configurations.

In this paper, a new category of the wheel-leg hybrid robot is presented. The proposed mechanism can compose large wheel diameter compared with the previous hybrid robot to realize a greater ability to climb obstacles. A prototype model of one Wheel-Leg module of the proposed robot mechanism has been developed to illustrate the concept. Actual design and mode changing experiment with a test mechanical module is also presented. Basic movement tests and a test of the basic properties of the rotational fingertip are also shown. The Basic configurations of wheel-leg retractable is considered well. The integrated mode is also described.

Underwater robot using piezoelectric fiber composite with the merits of being soft and capable of large displacement, is proposed in this paper. By pasting the thin and flexible composite on to a structure, meandering motion observed in underwater creatures such as fish can be realized by generating propagation wave for propulsive force. Hence, a compact, effective and powerful robot capable of underwater movement can be realized. This paper describes the mechanism design and control method of the robot using piezoelectric fiber composite based on the principle of underwater creatures and shows the feasibility of the proposal by the experimental results of two developed prototypes.

Even with the eyes closed, humans are able to grip an object with minimal force without such information as the coefficient of friction or the weight. Tactile sensors capable of detecting slippage are necessary for this gripping action to be realized in a robot hand. Heretofore, many slip sensors were developed and produced, but there was not a slip sensor of simple structure practical for installation on fingertips of a robot hand. Therefore, we propose a low-profile/lightweight slip sensor of simple structure. The special properties of pressure conductive rubber are utilized as a detection device in this sensor. In this paper, we discuss the results of trial manufacture and of slip detection property testing of this sensor. Moreover, we will report the results of slip prevention experiments by this prototype slip sensor, and indicate that the pressure conductive rubber is promising as a material of slip sensor.

A new golf swing robot performing high-speed golf swings has been developed by the authors. This paper deals with the motion planning problem of the golf swing. At first, the reverse time symmetry and single pendulum inherences in manipulator dynamics are introduced. Then, by utilizing the first inherence, the Rest-to-Point motion planning problem for the backswing and downswing is transformed into a reverse time symmetric Point-to-Rest motion planning problem. Finally, based on the second inherence, a Proportional plus Gravity and Coupling Torque Compensation (PGCTC) control scheme is developed to solve both the reverse time symmetric Point-to-Rest motion planning problem for the backswing and downswing and the forward time Point-to-Rest problem for the follow-through. Simulation shows the effect of the proposed method.

This paper reports a motion planning scheme for a high performance robot aiming to realize the motion control skills exhibited by professional golfers. The robot has a dexterous mechanism with similar distribution of actuators' capability and a pair of mechanical joint stops like human beings. The proposed motion planning method combines target dynamics together with port-controlled Hamiltonian (PCH) system theory resulting in an energy controller which not only takes the mechanical joint stops into account but also realizes torque compensation from a high-power actuator to a low-power actuator. Simulation and experimental results prove the proposed method can generate the golf swings with specified hitting speed and finish position for our specially designed robot.

In this paper, a tetrahedral mobile robot with central axis for transformation to the flat vehicle is presented. The throwable robot with the function of going into narrow spaces when its in the flat-vehicle mode is explained in detail. A prototype has been developed to illustrate the concept. Motion experiments confirm the novel properties of this mechanism: mode changing function and omnidirectional motion. Basic Motion experiments, with a test vehicle are also presented.

In this paper, the basic running performance of the cylindrical tracked vehicle with sideways mobility is presented. The crawler mechanism is of circular cross-section and has active rolling axes at the center of the circles. Conventional crawler mechanisms can support massive loads, but cannot produce sideways motion. Additionally, previous crawler edges sink undesirably on soft ground, particularly when the vehicle body is subject to a sideways tilt. The proposed design solves these drawbacks by adopting a circular cross-section crawler. A prototype. Basic motion experiments with confirm the novel properties of this mechanism: sideways motion and robustness against edge-sink.

To achieve a human like grasping with a multi- fingered robot hand, the grasping force should be controlled without using information from the grasped object such as its weight and friction coefficient. In this study, we propose a method for detecting the slip of a grasped object using the force output of Center of Pressure (CoP) tactile sensors. CoP sensors can measure the center position of a distributed load and the total load applied on the surface of the sensor, within 1 ms. These sensors are arranged on the fingers of the robot hand, and their effectiveness as slip detecting sensors is confirmed in tests of slip detection during grasping. Finally, we propose a method for controlling grasping force to resist tangential force applied to the grasped object using a feedback control system with the CoP sensor force output.

A tactile sensor is developed with the aim of covering a robot’s entire structure, while reducing wiring requirement and ensuring high-speed response. The sensor detects the center point of load distribution on 2-dimensional surfaces as well as the overall load. There are only 4 signal wires from the sensor. The sensor response time is nearly constant (within 1ms) regardless of the number of detection elements, their placements or sensor areas. In this paper, the principles behind the operation of this sensor and the results of experiments using the sensor are described. Further, a developed tactile element composed of pressure-sensitive materials and chip-resistors is described.

In this paper, adaptive robust force/motion control strategies are presented for mobile manipulators under both holonomic and nonholonomic constraints in the presence of uncertainties and disturbances. The proposed control is robust not only to parameter uncertainties such as mass variations but also to external ones such as disturbances. The stability of the closed-loop system and the boundedness of tracking errors are proved using Lyapunov stability synthesis. The proposed control strategies guarantee that the system motion converges to the desired manifold with prescribed performance and the bounded constraint force. Simulation results validate that the motion of the system converges to the desired trajectory, and the constraint force converges to the desired force.

Motion control is investigated for a nonholonomic mobile manipulator using hybrid joints, which could be switched to either active (actuated) or passive (underactuated) mode. The mobile platform is driven with two independent wheels and the arm has hybrid joints in the horizontal plane. The dynamic constraints are shown to be a first-order nonholonomic for the mobile base and a second-order nonholonomic for the link when one of hybrid joints is underactuated. Two Different motion control methods for mobile underactuated manipulator with one passive joint are proposed. Simulation and experimental studies show the effectiveness of the proposed methods

A sensor attached to the sole of the foot of a biped robot to detect a zero moment point (ZMP) is proposed and basic experimental results are presented. The sensor simultaneously detects the center position of a two-dimensional distributed load on the surface of the sensor and the total load of the distribution. The sensor is sheet-like in form, lightweight (0.2 g/cm2), offers a high-speed response (within 1 ms), and needs a little wiring (four wires). In the present paper, the principle of the above-described sensor and its characteristics are described, and the ZMP locus is shown for a walking biped robot with the attached sensors

A new golf swing robot performing high-speed golf swings has been developed by the authors. This paper deals with the motion planning problem of the golf swing. A target dynamics based control scheme is proposed. This control scheme maps the robot to a harmonic oscillator (target dynamical system). Based on the target system, energy control is adopted to realize a specified hitting speed at a specified impact position in the swing. After the swing passed through the impact position, PD control and PD plus gravity, joint torque and coupling torque compensation control are adopted to smoothly slow down and stop the swing at a specified finish position. Simulation and experimental results show the effect of the proposed method

a human-symbiotic safe mobile manipulator with nonholonomic constraint is described to realize human safety, impact force absorbing and task fulfillment. The robot consists of an arm covered with soft materials and hybrid joints, which can be put into active or passive mode as needed. In an unexpected or expected collision with human, the arising impulse force is attenuated effectively by the physical model formed with the hybrid joint and the soft material. Owing to the displacement movement of the link when the joint is passive, a recovery control algorithm has been developed for the end-effector to maintain its desired task position after the collision. Simulation and experiment results confirm that the proposed physical model is suitable for robot working in the human-robot symbiotic environment and the control method are useful for robot’s task fulfillment.

In this paper, a human-symbiotic robot is described to realize human safety, impact force absorbing and task fulfillment. The robot consists of an arm covered with soft materials and hybrid joints, which can be put into active or passive mode as needed. In an unexpected or expected collision with human, the arising impulse force is attenuated effectively by the physical model formed with the hybrid joint and the soft material. Owing to the displacement movement of the link when the joint is passive, a recovery control algorithm has been developed for the end-effector to maintain its desired task position after the collision. Simulation results confirm that the proposed physical model is suitable for robot working in the human-robot symbiotic environment and the control method is useful for robot's task fulfillment.

Usually there exist joint stops in a manipulator due to the mechanical structure of joints. In conventional manipulators, a collision between an arm and the structural joint stop is avoided by barriers from both software and electronic hardware. Comparing to these conventional manipulators, human utilizes structural joint stops usefully in many manipulations, especially in hyper dynamic manipulations such as golf swing. A basic concept and general method to utilize the structural joint stop in a manipulator, aiming at improving the dynamic capability of the manipulator in a large extent by simple and smart structure have been given in our previous work. To show the feasibility of the proposal, a hyper dynamic manipulator utilizing joint stops has been fabricated and fundamental experiments(see video) have been carried out. Experimental results have shown the important role of the joint stop, to realize an efficient hyper dynamic manipulation while in a smart structure. This paper describes the developed prototype and experimental results.

A sensing system using sonar ring for mobile robot capable of rapid detection and avoidance of obstacles has been developed. In the system, a method of firing all ultrasonic sensors on a sonar ring simultaneously is used and to eliminate error data due to crosstalks in the simultaneous firing, a filtering method based on pattern matching using neural network is used. To show the advantage of this method, a prototype of sensing system with 24 sonar sensors is fabricated. By installing the system to mobile robot, a scanning rate up to 50 Hz of panoramic detection is achieved. The experimental results showed that a rapid navigation of mobile robot in a real environment becomes possible. This paper describes the basic concept, the method for sensing and avoidance of obstacles, configuration of the developed sensor system and experimental results of obstacle detection and avoidance.

Usually the revolute joint of a manipulator cannot rotate all around, that is, in a range larger than 360 degree and it is limited by a mechanical element, the so called "joint stop". In conventional manipulators, a collision between an arm and the joint stop is avoided by barriers from both software and electronic hardware. Comparing to those conventional manipulators, human utilizes structural joint stop usefully in many manipulations, especially in hyper dynamic manipulations. In this paper, we propose a general utilization of structural joint stop in a manipulator to improve the capability of the manipulator in a large extent by a simple modification of the structure of joint. This paper gives the basic idea for utilizing structural joint stop in a manipulator, expected effects of the proposal, the method how to utilize the joint stop and simulation results.

Trajectory of an underactuated manipulator is usually generated according to both kinematics and dynamics of the manipulator, different to that of conventional manipulator. The real trajectory by a real system may differ greatly from the trajectory generated from the dynamics model because there always exist errors in the dynamic model, and the feedback control is less effective in an underactuated manipulator. A method for generation of optimal trajectory for the real system of an underactuated manipulator with nonholonomic constraints is proposed. By using this method, the dynamics model of a real system can be improved by learning, and an optimal trajectory is generated according to the model improved sequentially. The effectiveness of the method is confirmed by experiment with a golf swinging robot. The implementation and experimental results obtained of the control method are described.

To construct a practical system using a mobile manipulator, a new mobile manipulator with exchangeable active/passive joints has been developed by authors. In the mobile manipulator, passive joints play an important role in absorbing the mobile positioning error and reducing the internal force passively. By utilizing the exchangeable active/passive joints, a control algorithm. for automatic insertion of a long pipe based only on the angular information of the passive joints is proposed. By implementing the algorithm to the mobile manipulator, the insertion task was performed successfully. The proposed control algorithm is described and the experimental results are shown in this paper.

The motion control skill in human hyper dynamic actions, especially in ultra high speed actions, is being investigated by authors, toward a realization of hyper dynamic manipulation by robot. As a case study, a new golf swing robot to simulate human skill of motion control, that is, multistep acceleration by dynamic coupling drive, has been proposed by us. For simplicity, a prototype consisting of one actuated joint by a direct drive motor and one passive joint with a mechanical stopper has been developed. The purpose of this research is to realize smooth control of the robot while hitting a ball. The basic method and experimental results for controlling the motion of the robot in the case of hitting a ball are described in this paper.

A new golf swing robot to simulate the human skill of motion control in a golf swing has been developed by authors. It consists of one actuated joint and one passive joint with mechanical stopper. Optimal motion planning of the robot to satisfy the boundary conditions and non-holonomic constraint introduced by the stopper under different cost functions is discussed. A general method for motion planning of such types of system is described and the results of motion planning for the golf swing robot are shown.