Soft robotic hands are reliable for grasping objects of various shapes. However, they perform poorly in the high-precision manipulation of grasped objects because the modeling and sensing of soft actuator deformation are complex. To overcome this problem, in a previous study, we proposed a robust visual feedback control method for precise in-hand manipulation using parallel soft actuators. This method enables precise in-hand manipulation without measuring the soft actuator deformations. Generally, in the feedback control of a parallel drive system, the actuator force is converted into the force/torque applied to the grasped object. The conversion matrix constantly changes depending on the position/orientation of the object and the contact points of the soft actuators with the object. Consequently, accurately measuring the conversion matrix is difficult. Therefore, we estimated it as a constant matrix in our previous studies, and its robustness was confirmed experimentally. However, its theoretical robustness was not analyzed sufficiently. Therefore, in this article, we discuss the robustness of the estimated constant matrix through a mathematical stability proof. Furthermore, we investigate the characteristics of the estimated drive matrix. Then, we perform a numerical robustness analysis. Finally, the robustness of the proposed method is studied via the above investigation and verification experiments.

In this study, we investigate the effectiveness of haptic shared control (HSC) for underwater vehicle operation. Underwater vehicles such as remotely operated vehicles (ROVs) and autonomous underwater vehicles are widely used for underwater maintenance and inspection. In general, ROV operators need sophisticated skills as they have to control the position and orientation of an ROV using only limited information. To improve the path tracking performance of an ROV and support its operators, we propose applying HSC by incorporating haptic feedback in a control device. We conducted a pilot experiment in a water tank using an ROV. Subjects were instructed to operate the ROV from a start line to a goal line along a desired path while watching the video image sent from the ROV camera. They were also instructed to answer questions about the colors of objects on the desired path while controlling the ROV. The results suggested that the HSC decreased their mental workload, and increased their subjective feelings of concentration on the monitor and maneuverability of the ROV while no significant difference was found in the control performance.

A pneumatically-driven flexible finger equipped with a flexible sensor is realized for improving the performance of the soft robotic hand. First, we propose a flexible angle estimation sensor. This sensor measures the change in the amount of light passing through polyurethane material and estimates the angle with high repeatability. Next, we design a flexible finger that makes this sensor easy to incorporate. The flexible fingers are produced with a multi-material 3D printer that can use flexible material. The flexible finger can accommodate the proposed flexible sensor within it. It is possible to place the sensor's signal line in the air pressure pipeline. Because the flexible finger is produced with a 3D printer, variations in each model's characteristics are small as compared with manufacturing through molding. In this paper, we show an improvement of positional accuracy in the proposed flexible finger using angle feedback control from the proposed sensor. The effectiveness of this sensor is also shown to solve the problem of vibration problems for the flexible finger during high speed motion.

Safety and acceptability are critical issues when people are interacting with robots in healthcare. Traditional robot arms are hard and inflexible, and may cause harm to users on impact. Soft robotic arms may be safer and more acceptable in these situations. Similarly robot fingers made of soft materials may be more acceptable and safer than fingers made of hard materials. Robot designers need to know how best to design arms for healthcare scenarios. There is limited research on the acceptability of soft robotic arms and fingers to date. This study aimed to investigate people's reactions to the touch of soft robotic arms and fingers, compared to more traditional hard forms, and to human arms and fingers. A second aim was to investigate people's perceptions of the usefulness of the arms and fingers for healthcare tasks. Thirty five community participants were blindfolded and participated in touching tasks for: 3 arms (soft robot, hard robot, and human) and four fingers (soft robot, medium robot, hard robot, and human) in a randomised order. The soft arm was rated significantly more human-like but also more fragile and less reliable than the hard arm. Participants perceived the soft arm as good for intimate tasks like washing the body, but the hard arm was perceived as better for weight-bearing tasks. The soft finger was rated significantly more creepy, fragile and unreliable than the other fingers. The medium robot finger was rated the most human-like of the robot fingers and was the favourite robot finger. These findings suggest people perceive soft robots to be more fragile than hard robots and as more appropriate for personal tasks. Overly soft fingers may be too creepy to be acceptable.

This paper proposes an actuation mechanism of robotic arms using rotational counterweights. The robotic arm is driven by gravitational and inertia forces of the counterweight mounted on links. The joints of the robotic arms are able to rotate freely though this robot does not have any actuators on joints. Thereby, the robot has high joint flexibility. Additionally, the counterweight system reduces the moment applied to base link caused by gravity effect since the posture of the arm keeps a gravitational equilibrium state statically. Therefore, it is expected that the robotic arm is utilized in narrow and high altitude environment or on small vehicles. This paper also proposes a control method of the new robotic arm. The control method realizes precise position control of joint angle. The usefulness of the method is demonstrated by numerical simulations and experiments using a prototype.

This paper presents an inflatable robotic arm controlled by a joystick to be used for healthcare applications. The arm is constructed almost entirely of plastic elements: inflatable links, air bag actuators, and acrylonitrile butadiene styrene (ABS) joints. Therefore, it is softer and lighter than typical robotic arms that are made of metal and heavy elements. Because the softness and lightness of the inflatable robotic arm is intrinsically safer, it is suitable for healthcare applications. In this paper, a new control method is proposed which allows the inflatable system to be controlled with a joystick. To verify the usefulness of the proposed method, we used an inflatable robotic arm with four degrees of freedom (4 DOF) to obtain experimental results for the control performance of the inflatable robotic arm. Moreover, we conducted preliminary tests which simulated patients controlling the robotic arm with a joystick in order to assist with eating their meals.

This paper focuses on a possibility of application of a method to simplification of robotic arm systems by removal joint encoders. The method which has been proposed achieves precise control of robotic arms using visual feedback with camera calibration errors. In this paper, the method is improved for an encoderless robotic arm system to ease several difficulties to use encoders. Encoders are fundamental components of most robotic arm systems to control joint angles precisely. However, encoders sometimes make some difficulties of wiring, structure and maintenance. The enhanced method for encoderless robotic arm systems is able to overcome these difficulties. Moreover, this method is also useful for robotic systems which have encoders since the method is able to be activated as a fail-safe system of robots with encoders just in case encoders are broken down. This new approach acquires information on joint angles from not encoders but estimation using back electromotive force. The information on joint angles includes some errors. The method, however, realizes precise manipulation. It is finally shown that this method allows errors caused by not only from camera calibrations but also joint angles.

This paper presents a motion control method for multi-joint robots, which involves the use of mechanical elastic elements with adjustable stiffness. Elastic elements are installed at each joint of the robots, and the proposed control method adaptively adjusts the stiffness of the elements. This stiffness adaptation minimizes actuator torque, while the control method generates fixed periodic motions. Advantages of the control method are that it works well without requiring the knowledge of the exact values of physical parameters or elaborate numerical calculations. We mathematically show trajectory-tracking performance of the control system, and the effect of torque minimization. Experimental results demonstrate that the proposed control method greatly reduces actuator torque while generating the desired motions.

This paper proposes an energy efficient method for pick and place tasks of SCARA robots. In the proposed method, an adaptive elastic device at each joint of a robot is effectively utilized to reduce the total energy of pick and place tasks. For practical pick and place tasks, start/end points must be changed depending on the required task. In the previous works related to adaptive elastic device for energy saving, it is not clear how to change the start/end points. In this paper, a method to change positions of start/end points of pick and place tasks is proposed. The energy efficient performance of the method is demonstrated by the experiments, in which a SCARA robot with an air cylinder and a vacuum pad is used for the pick and place task of chocolate plates.

This paper proposes a new feedback control method based on simultaneous use of position/image sensors for a robot arm. The proposed method can guarantee the stability of the robot motion with the position sensors and can enhance the reliability. Moreover, the image sensors can achieve fine positioning without being affected by position sensors. The proposed control method needs neither kinematic calibration nor camera calibration. The effectiveness of the proposed method is demonstrated by experimental results in which a robot with 3DOF and a stereo camera set are utilized.

This paper examined the effects of the nonlinear mechanical properties that originate in the subcutaneous geometric structure and the subcutaneous mechanoreceptors of the fingertip. We briefly analyzed nonlinear biomechanical properties of the cutaneous surface and showed that the subcutaneous mechanoreceptors affect tactile perception as nonlinear transducers. The validity of our hypothesis was assessed by experimentally investigating the vibrotactile sensation of a fingertip.

Estimation of joint torque is an important objective in the analyses of human motion. In particular, many applications seek to discern torque during a desired human motion, which is equivalent to solving the inverse dynamics. The computed torque method is a conventional means of calculating inverse dynamics. The obtained torque, however, invariably includes errors resulting from inexact inertial and viscoelastic parameters. This paper presents a method for solving the inverse dynamics of a human arm passively during tracking. To achieve precise human motion tracking, iterative learning control is used for motion generation. Some experiments that target a human arm are executed to validate the proposed method.

In this paper, we propose a pitch and roll control system based on a float-shift mechanism for small underwater robots. The proposed system presented here takes advantage of the restoring moment generated by the gravity force and the buoyancy force. The proposed system can maintain the attitude of a robot in its desired orientation. In previous works, we developed a prototype system that only controls the pitch angle of a robot. This paper extends the control system to provide roll and pitch control. In this paper, we explain the static model of the proposed system and how to control the pitch and roll angles of an underwater robot. Next, we conduct numerical analysis in order to investigate the performance of the proposed system. Lastly, we show that the proposed system is able to change the pitch and roll angles of an underwater robot though several experiments in a test tank.

In this paper, we present the design and development of an attitude control system for a human-sized ROV (Remotely Operated Vehicle) with a dual-manipulator system. The main operations of the ROV are sampling operations for biological researches, geological researches and archaeological explorations in Lake Biwa, the biggest lake in Japan. In order to achieve these operations, we propose an attitude control system and explain the availability of the proposed system in this paper. For the design of the attitude control system, simulation analysis was conducted. Based on the analysis, a pitch angle control system was designed and developed as the prototype of the attitude control system. An experiment was conducted in a test tank to demonstrate the effectiveness of the proposed system. The result shows that the developed system is useful for pitch angle control of the ROV.

As described in this paper, we propose an attitude control system for underwater vehicle/manipulator systems (UVMSs) based on control of the position of the center of buoyancy with respect to the center of gravity. Control of the center of buoyancy is accomplished using movable float blocks. The attitude control system is useful to control the pitch angle of UVMSs to enhance their performance and to improve their efficiency of underwater operations. A UVMS that has two 5-degree-of-freedom (DOF) manipulators was developed to verify the effectiveness of the proposed attitude control system. This paper presents a numerical study and some experimental results obtained using the UVMS with the attitude control system. We experimentally confirmed that the proposed system can change the pitch angle of the vehicle between -120 and +105 deg. In another experiment, attitude-maintenance control was conducted. Results show that the proposed system can maintain the vehicle's horizontal attitude during motion of the manipulators.

This paper proposes a task space trajectory tracking controller based on resonance for multi-joint robots. This controller generates desired motions, which are specified in the task space, while adjusting stiffness of mechanical elastic elements installed in each joint of the robots. This controller also adjusts equilibrium angles of the elastic elements. These parameter adjustments minimize actuator torque. Advantages of the proposed controller are to work without using exact parameter values of the controlled systems nor huge numerical calculations. We mathematically discuss stability of the controlled systems. Simulation results demonstrate the effectiveness of the proposed controller.

The basis-motion torque composition (BMC), which generates feedforward torque for precise joint-trajectory tracking of a multi-joint robot arm, has been recently suggested. It is based on four arithmetical operations of time-series torque data of several motions preliminarily obtained by the iterative learning control. The remarkable advantage of BMC is to generate desired feedforward torque without prior information of dynamics parameters. However, the class of torque generated by the BMC has been restricted. The paper presents an enhanced BMC which permits generation of motions with different motion-velocity profiles among joints of a multi-joint robot. The algorithm is presented and the validity of algorithm is confirmed through numerical simulations in the case of a two-joint robot arm under the influence of gravity. Furthermore, a class of applicable systems of BMC approach is discussed.

We describe here the development a joint mechanism for underwater robotic manipulators. Arms of underwater robots require small-scale bodies and high waterproofing properties. In most of underwater robots, electric motors are used as actuators to drive the robotic arm/arms, but using electric motors for underwater manipulators may be problematic due to the size/weight of the robotic arm and need to waterproof the electric motors. We develop a joint mechanism composed of combinations of rigid and flexible members, which can be deformed by a prismatic actuator fixed onto two rigid parts. We utilize a leaf spring as the flexible joint and a McKibben actuator driven by water hydraulic pressure as the prismatic actuator. The number of members in this mechanism is smaller than that of a mechanism composed of a combination of one pulley and one coil spring. One advantage of this mechanism is the avoidance of gears, thus eliminating sliding parts from the joint.

This paper proposes passivity-based controllers for periodic motions of multi-joint robots with impact phenomena. Even the robot motions with impact phenomena have complex dynamics, we try to analyze stability of the controlled systems by adopting some assumptions and using Lyapunov-like functions. At first, we present a passivity-based feedback controller. Secondly, we present a repetitive controller based on passivity-based iterative learning control. These controllers generate desired periodic motions, which are specified by users of the controllers. Advantages of the proposed controllers are to works well without using exact parameter values of the controlled systems nor huge numerical calculations.

In this paper, we propose a new mechanical structure for adjustable stiffness devices with lightweight and small size. The proposed structure utilize a ball screw mechanism to adjust a relationship between infinitesimal displacements of joint rotation and a linear spring. Then, stiffness around the joint is adjusted. Unlike many of other adjustable stiffness structures, available elastic energy of the elastic element is maximum when the stiffness of the proposed structure is maximum. Therefore, the elastic element of this structure can be smaller and more lightweight than the other structures. Another advantage of the proposed structure is to require fewer and smaller mechanical parts, because the proposed mechanism mostly requires the ball screw mechanism and the linear spring. We developed an actual hardware to test the proposed structure.

This paper proposes a new trajectory tracking control method for multi-joint robots by combining stiffness adaptation and iterative learning control. The proposed controller achieves trajectory tracking while optimizing stiffness of elastic elements installed in each joint of the robots. Even though the multi-joint robots have nonlinear dynamics and multi degree-of-freedom, the stiffness optimization realizes high energy efficiency as if we utilized resonance. An advantage of the proposed control is to work well without using exact parameter values of the robots. Since it seems that adaptive control and iterative learning control have not been used simultaneously, this paper newly proposes a methodology to appropriately combine stiffness adaptation and iterative learning control. This combination enables trajectory tracking and convergence of the stiffness to the optimal one. These properties can not be achieved by our previous controllers.

This paper focuses on generation of feedforward torque for precise joint-trajectory tracking of a multi-joint robot arm with serially connected links. A proposed method called basis-motion torque composition, based on four arithmetical operations of time-series torque data for several motions, allows to generate feedforward torque for a motion whose final posture and time profile are specified. The torque-generation algorithm is presented, and the experimental results by a two-joint robot arm are illustrated. The experimental results demonstrate that the tracking errors of angular velocities by the basis-motion torque composition tend to be smaller than those by the computed torque method.

In this paper, we apply an energy saving control method to a simulation model that includes some dynamics of biped walking robots. This control method was proposed by the authors, and is based on resonance of multi-joint robots. An advantage of the control method is to work well without using exact parameter values of the controlled systems nor huge numerical calculations. This paper discusses how to apply the proposed control method to walking motions as a first step to realize energy saving biped robots. For this purpose, we consider some dynamics of the biped robots. Simulation results showed that the proposed control method can generate energy saving walking motions.

This paper focuses on the Riemannian distance and its application to skilled-motion plannings for the system. The Riemannian distance from one pose to another and vice versa is defined as the minimum curve-length measured by the Riemannian metric based upon the system inertia matrix among all curves connecting the two poses. The minimum-length curve in this meaning is called “geodesic” and reflects a movement of the system affected only by inertia-tensor-originated force (i.e., pure inertia, centrifugal, and Coriolis forces). In order to investigate in detail such a movement along the geodesic, some computer simulations are conducted in the cases of planar motions by a 4-DOF robot arm and biped walkings by a wholebody robot. It is shown through simulation results that movements attaining the Riemannian distance (natural movements in inertial actions) in the two cases tend to be similar to those in human skilled motions when human-scale robot models are chosen. Based upon the Riemannian distance, motion plannings for multi-body systems using physical properties inherent in their own physical structures are discussed.

This paper proposes an energy saving control method of periodic motions composed of multi-frequency components based on adaptive stiffness optimization for mechanical systems. In the case of sinusoidal motions, the concept of resonance can be easily applied to determine an optimal stiffness. This paper tries to extend resonance to deal with periodic motions composed of multi-frequency components. For this purpose, we define a cost function, which represents an amount of actuator torque. Next, an optimal stiffness is defined as the stiffness that minimizes the cost function. The relationship between the optimal stiffness and the desired motions is analytically derived. Based on the preparations, we propose two types of controller. These controllers generate the desired motions using the least amount of actuator torque as possible by optimizing the stiffness adaptively. The controllers work well without using parameters of the objective systems. Stability and effectiveness of the stiffness optimization are proved mathematically. Simulation results demonstrate the effectiveness of the proposed method.

This paper proposes an iterative learning control (ILC) scheme for a class of redundant robot arms to acquire the desired control input signals that produce an endpoint trajectory specified in task space. The learning update law of control input signals is constructed only in task space by modifying the previous control input through adding linearly an endpoint velocity trajectory error. Although the dimension of the task space is strictly less than the DOF (degrees-of-freedom) of the robot arm, the proposed method need neither consider any inverse kinematics problem nor introduce any cost function to be optimized and to determine the inverse kinematics (or dynamics) uniquely. Convergence of trajectory trackings to the specified one is shown by numerical simulations in both cases 1) free-endpoint motion and 2) constraint-endpoint motion with specified contact force. A theoretical proof of convergences in the case of free-endpoint motion is given on the basis of an approximated dynamics linearized around a desired solution in joint state space.

This paper proposes a tracking control method of sinusoidal motions utilizing stiffness adjustment of mechanical elastic elements for serial link systems. Although dynamics of the controlled objects is nonlinear, the stiffness adjustment realize a condition similar to a resonance of linear systems. We present a controller that adjusts stiffness of the elastic elements to reduce torque requirement of actuators while generating desired motions. The proposed controller works without using dynamics models nor parameters of the controlled objects. Stability of the controller is proved, and tracking errors are guaranteed to converge to a certain region. Simulation results demonstrate the validity of the proposed method. We also present an application of the proposed method to power assist systems.

In this paper, we propose a power assist system that amplifies sinusoidal human's torque and attains minimization of control input requirement using an impedance control and resonance. This impedance control is designed to realize the following features: 1) sinusoidal torque amplification. 2) Minimization of control input requirement by adjusting stiffness. 3) No requirement for the knowledge of amplitude, frequency and phase of human's torque. 4) No requirement for myoelectric signals. 5) Satisfaction of causality. Convergence of the proposed controller is proven theoretically. Simulation results verify the validity of the control scheme including some conditions not discussed in the theory. Experimental results show the validity of the control system including real human's complex dynamics and feedback loops

In the case of control for nonlinear mechanical systems, which have complex actuator dynamics, the total system becomes high-order and nonlinear. Generally speaking, it is not easy to tune feedback gains for state feedback. In this paper, we propose a feedback control scheme for motion control of nonlinear high-order systems. We prove that the proposed scheme can improve the trajectory tracking performance of a robot manipulator system with actuator dynamics. Moreover, some simulation results demonstrate the effectiveness of the proposed control scheme

This paper proposes power assist controllers that realize control input minimization using resonance of passive elements. To satisfy the objective, two types of systems are considered and controllers are proposed for each system. These controllers require no biosignals and satisfy causal manner. Convergence of the controllers is discussed theoretically. Simulation results show the validity of the proposed controllers. Experimental results show effectiveness of the proposed controller when the systems are used by a human operator

This paper presents a new ankle-foot orthosis (AFO) for hemiplegic patients. The proposed AFO has a pneumatic passive element with a rotation axis at the ankle joint. The ankle-joint constraint force can be adjusted according to the subject's gait phase. We confirmed that adjusting of the pneumatic pressure in the proposed pneumatic passive element enables the device to duplicate the performance of many conventional AFOs

Multiple links structure/wire driven (tendon) system is one of robot structures, that has a serial-link structure driven by a wire-drive mechanism. It can be applied to a manipulator like elephant's trunk, a moray arm, the backbone of a humanoid, etc.. However, transformations among coordinates (task-oriented, joint-angles and wire-length) are very complicated. Therefore, it is not clear how sensory feedback control laws should be realized for the system. We propose a sensor-actuator map in which the sensor coordinates and actuator coordinates are arranged. By finding the paths from the sensor coordinates to the actuator coordinates on the map, the sensory feedback control laws can be built.

In this paper, three fundamental techniques to improve laminated passive elements are proposed for wearable haptic displays. First, a new shape of sheet surfaces in laminated passive elements is introduced. Sliding motion between laminated sheets causes unexpected bad effect to static stiffness. An effective triangle shape for sheet surfaces, which prevents from sliding, has been analyzed. This technique improves the stiffness of laminated elements. Next, a binary stiffness control method is proposed. It realizes arbitrary stiffness in a certain resolution only by on/off air valves. A parameter designing strategy for linear stiffness change is also described. Last, a technique to realize controllable multi-dof constraints is introduced. It realizes a multi-dof module of laminated elements, so it is an essential technique for multi-dof haptic devices. The validity of all the proposed techniques has been verified by some preliminary experiments.

Most research so far in robot control has assumed either kinematics or Jacobian matrix of the robots from joint space to Cartesian space is known exactly. Unfortunately, no physical parameters can be derived exactly. In addition, when the robot picks up objects of uncertain lengths, orientations, or gripping points, the overall kinematics from the robot's base to the tip of the object becomes uncertain and changes according to different tasks. Consequently, it is unknown whether stability of the robot could be guaranteed in the presence of uncertain kinematics. In order to overcome these drawbacks, in this paper, we propose simple feedback control laws for setpoint control without exact knowledge of kinematics, Jacobian matrix, and dynamics. Lyapunov functions are presented for stability analysis of feedback control problem with uncertain kinematics. We shall show that the end-effector's position converges to a desired position in a finite task space even when the kinematics and Jacobian matrix are uncertain. Experimental results are presented to illustrate the performance of the proposed controllers.

A vision-based approach to the motion control of pneumatic group actuators is presented. Sensing of plate locations consisting of the pneumatic group actuator is essential to cope with the variation of elastic tubes of the actuator. Plate locations are measured by a vision system and the motion of the actuator is controlled using a roughly identified relationship between air pressure imposed on individual tubes and the plate locations. Proposed motion control is evaluated experimentally.

A new method to analyze the dynamics of underwater robot manipulators is proposed in this paper. In the proposed method, hydrodynamic terms such as added mass, drag and buoyancy in dynamics of underwater robots are obtained by iterative learning control and time-scale transformation. The advantage of the proposed method is not to use parameter estimation of the dynamics. In this paper, we explain that the proposed method can be applied to hardware design, motion control and motion planning of underwater robots. Moreover, the experimental results using a 1-DOF and a 3-DOF manipulator demonstrate the effectiveness of the proposed method.

A new type of passive element with variable mechanical impedance is proposed. Since the proposed elements are lighter, smaller and softer than the previous passive elements, these elements have the possibility to develop wearable robots. In the paper, the principles of the variable mechanical impedance are explained and the basic experimental results demonstrate the performance of the proposed element. Moreover, a wearable robot for virtual reality is developed by using the proposed element and it is shown that the developed wearable robot is utilized as a force display in the virtual world.

A type of robot system for a force display which has a serial link structure driven by a parallel mechanism is proposed. This system consists of serially jointed links as usual industrial robots and parallel wires that are driven by actuator units located on the base. Therefore, there are advantages of both a serial link structure robot and a parallel wire driven robot, such as large work space, high speed, safety, and so on. In previous analysis of the parallel wire driven mechanism, the object is a single rigid body, so that the driving principle in such cases was obvious based on a theory of "vector closure". However, in the case of the proposed system, the driving principle has not been cleared, because the mechanism is very complicated. In this paper, we show static conditions for each joint of the proposed system based on partial vector closure corresponding to the serial link structure/parallel wire system. Moreover, a prototype of the force display system is developed by using the conditions. Finally, the usefulness of the proposed system is demonstrated through some experimental results.

Conventional robot systems consist of hard mechanical elements and actuators for exerting a force to objects efficiently. These systems are taking remarkable activities in industrial field but not in daily life of human. Main disadvantage of these robot systems is the fatal risk to human body when they lose control in the case of breakdown. On the contrary, we focus on flexible robot systems that work passively without actuators. These systems are intrinsically safe to human body since they never exert any excessive forces. In this paper, we present a novel passive element that varies its stiffness to internal vacuum pressure. This passive element is a soft vinyl tube embedded with Styrofoam beads, and is therefore soft and light contrary to conventional passive elements. Using the flexibility and the controllable stiffness, we developed a wearable haptic display that provides a sensation of elasticity and viscosity. The design of this new passive element and the applications to haptic displays are discussed.

Most researches so far on robot control have assumed that the exact kinematics and Jacobian matrix of the manipulator from joint space to Cartesian space are known. Unfortunately, no physical parameters could be derived exactly. In addition, the robot is required to interact with its environment and hence the overall parameters would change according to different tasks. In the paper, simple feedback control laws are proposed for setpoint control of robots with uncertain kinematics and dynamics. We show that the end-effector's position converges to a desired position in a finite task space even when the kinematics is uncertain.

We develop pneumatic actuators composed of multiple elastic tubes. Elastic tubes with mechanical constraints have the capability of performing a single various motion while multiple motions cannot be performed. Actuators composed of single-motion tubes can realize multiple motions by controlling air pressure imposed on individual tubes. We present the concept of group actuators and develop their prototypes.

Robust manipulation strategies of deformable objects is presented. Manipulation of deformable objects can be found in many fields such as the garment industry and food industry. Guidance of multiple points on a deformable object is a primitive operation in the manipulation of deformable objects. In this guidance, the points often cannot be manipulated directly. A model of the manipulated deformable object is needed in order to perform these operations. It is, however, difficult to build a precise model of a deformable object. Thus, we need a robust control scheme that allows us to realize the operations successfully despite discrepancy between a manipulated deformable object and its model. We firstly derive a mathematical model of deformable objects for their manipulation. Second, indirect simultaneous positioning operations of deformable objects are formulated. Then, we propose a PID feedback control law with the rough object model to realize the manipulation. Furthermore, we propose a simple PID feedback control law without deformation model. The validity and the robustness of the proposed manipulation method is shown through simulation results.

Behavior of deformable cylindrical actuators is qualitatively investigated with respect to constraint topology. Elastic shells expanded by air pressure have a capability of generating various motion directly by imposing mechanical constraints on the shells. In this paper, we propose a qualitative analysis of the motion of deformable cylindrical actuators with mechanical constraints.

Pneumatic motors have many advantages over electric motors, conventionally used for robotics and automation applications. However, the control performance of pneumatic motors is quite unsatisfactory, mainly due to the compressibility of air. We have recently proposed the development of hybrid pneumatic/electric motors where a small electric motor is used in conjunction with vane-type pneumatic motors, so that the hybrid actuator has better damping and ease of control than the pneumatic motor. In the present work, we consider the switching control of the hybrid actuator system, and, propose control algorithms for position, and trajectory control of the system. The performance improvement of the new actuator is illustrated by results of experimental implementation.

In general, it is difficult to present haptic information because arbitrary curvatures of the curved surface and many degrees of freedom need to be realized. On the other hand, psychophysical studies have suggested that human visual and tactile sensations have an illusory fusion characteristic. This means that we can recognize curved surfaces of objects through visual and tactile sensations, even if exact tactile information is not presented. Hence, by utilizing the human characteristic of sensory fusion, realization of a haptic device of curved surface can be simplified. From such motivation, the human fusion characteristics of visual and tactile sensation are measured, and are quantitatively analyzed. Based on the analyzed results, a haptic device of two-dimensional curved surface for three fingers is developed. In the haptic device, only four curved patterns are utilized instead of presenting many curved patterns. Performance of the developed haptic device is proved through evaluated experiments.

Most research so far on robot control assumes that the kinematics and Jacobian matrix of the manipulator from joint space to task space are known exactly. This assumption leads to several open problems in the literature of robot control and limits the potential research and applications of robots. In this paper, we present an approximate Jacobian PID control law for set-point control of robot with uncertain kinematics from joint space to task space. Simulation results are presented to illustrate the results.

A new object sorting device using impulsive manipulation and air floating is developed. First, two methods for impulsive object sorting are evaluated using computer simulation for the design of a sorting device. Second, an object sorting system with air floating is developed. Finally, the developed system is evaluated experimentally.

Pneumatic actuators are among a variety of actuators under active study by the robotics community, for the advantages they offer over conventional electric actuators widely used in present-day robots. So far, however, only pneumatic cylinders which provide linear actuation have been mainly studied in literature. Air motors are the pneumatic counterparts of electric motors, providing rotary actuation. They have several advantages over electric motors, especially high power-to-weight and power-to-size ratios, good compliance, and resistance to environmental hazards. We consider the model-based control of vane-type air motors which are the most commonly used class of air motors. Various control problems are considered namely, position control, trajectory control, adaptive control, and force control. Results of experimental implementation on an industrial vane-type air motor illustrate the effectiveness of the air motor as a high-performance rotary actuator.

A control method for positioning operations of deformable objects is presented. In many manipulative operations of deformable objects, it is required to guide multiple points on on object simultaneously. Moreover, the points often cannot be manipulated directly. A model of the manipulated deformable object is essential to perform these operations. It is, however, difficult to build a precise model of a deformable object. Thus, we need a new control scheme that allows us to perform the operations successfully despite of discrepancy between a manipulated deformable object and its model. We introduce a coarse model of a deformable object and develop a control law for its positioning operation. First, we propose a mathematical model of deformable objects for their positioning operations. Second, indirect simultaneous positioning operations of deformable objects is formulated. Then, we propose an iterative control method to realize a given positioning operation. The validity of the proposed method and the effect of model errors on the operation are examined through experiments using textile fabrics. Experimental results show that coarse model of deformable objects is effective for their positioning operations. Finally, we discuss the locations of robotic fingers, which apply forces to a manipulated object.

This paper proposes an iterative learning control scheme for impedance control of robotic tasks when the tool endpoint covered by soft and deformable material presses a rigid object or environment at a prescribed periodic force pattern. To this end, an iterative learning control scheme for a class of linear dynamical systems with a negative feedback structure is analyzed and convergence of the proposed learning update law after a sufficient number of repetitions is proved. It is shown that this convergence realizes impedance matching in a sense of electric circuit theory of the feedback system can be expressed as a lumped-parameter electric circuit. The iterative learning control scheme is then applied for a case of impedance control of robotic tasks when the characteristics of reproducing force of the deformable material is nonlinear in its displacement and unknown and the tool mass is uncertain. Simulation results are also presented, which show effectiveness of the proposed learning control scheme.

This paper aims to generalize the concept of impedance matching to cope with nonlinear dynamics that govern robotic systems with or without constraints. In the case of single degree of freedom dynamics the impedance matching is first introduced by the concept of balancing between load impedance and internal impedance. Generalization of this concept to the case of multiple degrees of freedom robots can be fulfilled implicitly via a coordinates transformation by constructing a negative feedback connection of two hyper-stable blocks (one is a parameter estimator and positioning and the other is a position and force regulator).

Most research on multifingered robot control has assumed that the Jacobian matrices from joint space to task space is exactly known. This implies that the locations of contact points, geometry of the object, kinematics of the multifingered robot hands must be exactly known. In this paper, a task-space feedback control problem of multifingered robot hands with uncertain Jacobian matrices is formulated and solved. The stability and robustness of the proposed controllers to the uncertainties in Jacobian matrices are analyzed.

Research of robotics aims to realise some aspects of human functions into mechanical system. Human can manipulate things skilfully without the exact knowledge of both dynamics and kinematics of arms. Our arms are also able to overcome singular position by moving along it or passing through it. The exploration of a robot controller to cope with the uncertainties in both dynamics and kinematics is an important step towards understanding the dextrous movement of mechanical systems. In this paper, simple feedback control laws are proposed for setpoint control of robots with uncertain kinematics and dynamics. We shall show that the end-effector's position converges to the desired position in a finite task space even when the actual Jacobian matrix is singular.

We developed a robotic orthosis capable of helping human forearm motion. This orthosis was carefully designed such that mechanical safety for human beings could be ensured. A power assist control scheme for the robotic orthosis is proposed. The performance of the developed system is examined by several experiments.

In general, it is difficult to present tactile information due to arbitrary curvatures of the curved surface and many degrees of freedom needed to be realized. On the other hand, psychophysical studies suggested that human visual and tactile sensations have an illusory fusion characteristic. It means that we can recognize curved surfaces of objects through visual and tactile sensations, even if exact tactile information is not presented. Hence, by utilizing the human characteristic of sensory fusion, realization of a tactile display can be simplified. From such motivation, the human fusion characteristics of visual and tactile sensation are measured, and are quantitatively analyzed. Based on the analyzed results, a tactile display is developed. In the tactile display, only four curved patterns are utilized instead of presenting many curved patterns. Performance of the developed tactile display is proved through evaluated experiments.

A novel control method for the indirect simultaneous positioning operation of soft deformable objects will be proposed. In some operations, multiple points on a deformable object should be positioned to desired regions simultaneously. In addition, these positioned points often cannot be operated directly and we have to perform these operations indirectly by controlling other points. We call these operations indirect simultaneous positioning operations. First, a simplified physical model of deformable textile fabrics is developed for their positioning operations. Second, the indirect simultaneous positioning operations are formulated using the proposed model. Based on the linearized model of the deformable objects, we will propose a novel control method for the indirect simultaneous positioning operations with visual sensors. In this method, positioned points can be guided to the desired locations by controlling the positions of operation points. It should be noted that our control method works well even if exact physical properties of fabrics cannot be given. Experimental results will show the validity of our proposed method. In addition, a condition to examine whether a given operation is feasible or not will be considered based on the linearized model.

We develop a virtual sports training system for "playing catch". In this system, a force display device using "parallel wire mechanism" is adopted to obtain performance suitable for sports training. We describe the detail of the device used in this system. Furthermore, some basic experimental results demonstrate that it is capable of achieving the high speed performance enough to implement sports training.

A new approach to the modeling of plain knitted fabrics is presented for their deformation control. There exist many manipulative operations that deal with deformable soft objects such as clothes, papers, wires etc. At present, most of these operations are still done manually by skilled workers. Therefore, it is important to automate this process, especially the process of knitted fabrics manipulation in the garment industries. However, automatic handling of knitted fabrics is difficult to perform by machines since the knitted fabrics are often deformed and extended during the operations. Furthermore, it is difficult to control the shape of the knitted fabrics. In order to control the deformation of knitted fabrics, a physical model of the fabrics is essential. In this paper, we present a physical model of plain knitted fabrics. Firstly, the structure of plain knitted fabrics is explained. Secondly, the behavior of cross points of a plain knitted fabric is investigated experimentally. Then, a physical model of the fabrics is developed based on the iterative structure of the fabrics. Model parameters are selected considering the behavior of the cross points. Finally, the validity of the proposed model is examined by comparing computed shape of a fabric with its real image.

In the case of mechanical contact between humans and robots including motion-support robots or robotic orthosis, safety is most important. Therefore, in the motion-support robot presented in this paper, hexahedron rubber actuators (HRA) which have light weight, small size and flexibility of motions are used. In this paper, the basic performance of an antagonized HRA unit which is actuated by air pressure is revealed at first. Moreover, effectiveness of a motion-support robot with 4 DOF is demonstrated through experimental results.

In many cases of human hand tasks, grasping and manipulation of objects are the most basic motions. However, the influence of physical characteristics of human finger on grasping motions have not yet been sufficiently clarified. For example, friction which is an important characteristic of grasping motions has hardly been investigated. In this study, we pay attention, to friction between a finger and an object. First, we develop a system to measure the friction on fingers. Next, based on the experimental results obtained from the measurement system, we propose a new model for the maximum static friction force on human fingers. Also, we find that the friction forces are changed depending on the directions of tangential forces. Finally, we realize a finger-tip which is made of silicone and whose characteristics are similar to those of human fingers.

In general, design and realization of hardware of haptic device for virtual reality are very difficult because of the complicated mechanisms required to realize many degrees of freedom. On the other hand, it is well known that humans have illusions on sensations such as vision. Therefore, it is expected that the hardware of virtual reality can be simplified if we make good use of human illusions. From such a motivation we measure human illusion on distances between the thumb and the forefinger and curvatures of the finger touch surface. From the measurement results, some important points are suggested to realize simplified haptic device.

High speed robots are an important component of modern assembly operations. In this paper, we describe the development of an ultrahigh speed robot named FALCON (Fast Load Conveyance), based on a wire driven parallel manipulation system. It achieves peak accelerations of up to 43G and maximum velocities of 13 m/s, even if considerably small DC motors (60W) are used. Due to the use of wires in actuation, the problem of vibration arises. We employ internal force control among wires to reduce the vibration by utilizing nonlinear elasticity from the wire mechanism. Experimental results on point to point control using the prototype system are presented under linear PD feedback, to illustrate the effectiveness of the system.

A time-scale interpolation for feedforward input torques of robot manipulators is proposed. In the proposed method, by interpolating input torque patterns obtained through learning control, another desired motion with a different time-scale or a different speed pattern can be precisely realized without the use of iterative operation for learning control. In simulation and experimental results, it is confirmed that an input torque pattern with arbitrary time-scale can be interpolated from four input torque patterns which have different time-scales.

A new type of pneumatic rubber actuators is proposed in this paper. The proposed actuator has a hexahedron structure which is suitable for making a rotational joint of a robot. We reveal the static characteristics of the actuator. It is experimentally demonstrated that a position feedback control law realize fast motions without oscillation. Moreover, we propose an adaptive gain method for the actuator to improve control performance. The effectiveness of the proposed method is confirmed through an experiment.

We explore possibility of virtual tennis in this paper. To realize virtual tennis, it is necessary to provide our hands with reaction forces from a tennis ball. Moreover, we need a head mounted display which is capable of showing moving objects with high speed. In this paper, we develop an equipment which makes reaction forces at a grip of a tennis racket. In the developed equipment, several wires are radially stretched from the grip in order to produce force and torque vectors with six degrees of freedom. The motion area of the proposed system is made clear at first. A novel method to determine wire tensions which is suitable for real time calculation is proposed. Finally, a basic experimental result is demonstrated.

The paper presents a sliding-mode based implementation of impedance control. The sliding surface is defined as a measure of the impedance error. When the system variables correspond to the sliding surface, then the desired impedance is implemented. The proposed control approach is composed of two parts. First, the system is controlled such that the state variables reach certain values corresponding to a neighborhood of the sliding surface. This part is performed using a control law consisting of a version of the computed torque and an impedance error-based switching function. The paper uses a new formulation of the acceleration signal for the computed torque calculation. The need for such a neighborhood is to avoid chattering which is typical to sliding-ode based control laws. The second part is performed using a PI-controller which ensures that the state variables further converge to some desired values corresponding to the sliding surface, such that the impedance error would tend to zero asymptotically. The proposed method has been verified experimentally. The method is practical and superior to other known methods.>

In the case of pneumatic bellows actuators, the external forces are easily and exactly measured from the pressure in the chamber of bellows because there is no sliding part which generates friction. Therefore, the bellows can be utilized as a force sensing actuator which works as a sensor and an actuator simultaneously. We design one joint of a robot which is activated by the two bellows actuators antagonistically. From some experimental results we disclose the static and dynamic characteristics of the proposed robot and confirm that it can move flexibly following external forces by making use of the force sensing ability.>

Describes the theory and an experiment for a fast path planning method that is available for moving obstacle avoidance for an autonomous mobile robot by the use of the Laplace potential. This new navigation function for path planning is feasible to guide, in real time, a mobile robot avoiding arbitrarily moving obstacles and reaching the goal. The experiment is conducted to verify the effectiveness of the navigation function in obstacle avoidance in the real world. The experimental systems are composed of a position sensing camera, a mobile robot and a computer for processing the position signal and for controlling the velocity vector of the robot. Three examples are presented; firstly the avoidance of a standing obstacle, secondly the avoidance of a moving obstacle in parallel lines-bounded space, and thirdly the avoidance of one moving obstacle and another standing obstacle. The robot can reach the goal after successfully avoiding the obstacles in all cases.

A new type of master robot for teleoperation is proposed. In the proposed master robot, several wires are stretched from a handle which is moved by a human operator. In this paper, the authors explain how many wires are necessary to control a handle with six degrees of freedom in a three dimensional space. Moreover, to feed forces from a slave robot to the handle, an algorithm to determine each wire tension is proposed. Finally, they show experimental results in which a handle is moved by seven wires in a three dimensional space.

A sliding mode based impedance controller is proposed. A continuous PI function is used near the sliding surface to avoid chattering as well as to reduce the steady-state error. Stability analysis is performed via Lyapunov theory. The proposed controller is realized on a DD-arm robot to test its performance. Its performance in the grinding process with different parameters of targeted impedance is also investigated. Through frequency spectrum analysis of the grinding force, criteria to choose the proper parameters of the targeted impedance for grinding tasks are suggested.>

A method is proposed to control the tip of a flexible multilink manipulator. The external sensory feedback controller can be easily constructed in task-oriented coordinates (for example Cartesian coordinates) without complex calculation for inverse kinematics or inverse dynamics. To realize the control method, a reduced-order approximate Jacobian which is formed by the motion of both rigid and flexible modes of the manipulator is introduced. The effectiveness of the proposed control method based on the approximate Jacobian is proved by using L/sub 2/ stability theory. The practical usefulness of the method is demonstrated by some experimental results for a flexible two-link manipulator.>

A control scheme for pneumatic drive systems is proposed. In general, static friction, Coulomb friction, and other nonlinear disturbances crucially worsen the performance of pneumatic drive systems. In the proposed control scheme, to guarantee robustness against such disturbances and stability of the motion, position feedback loops and pressure feedback loops are hierarchically connected. The stability of the motion is investigated by using a Lyapunov function. As a result, it is mathematically proven that the motion of pneumatic drive systems converges to a fixed desired point as time tends to infinity, even if the disturbances affect the motion. The effectiveness of this control scheme is demonstrated through some experimental results.>

A theorem is proposed and used to prove that a robot can track a desired trajectory described by time functions with required arbitrary accuracy, if a linear PD (proportional-derivative) feedback loop is provided at each joint of the robot and the velocity feedback gain can be set sufficiently large. It is shown how to apply the theorem to real robots in order to realize efficient, high-performance trajectory tracking. The relationship of the control to a linear PID feedback controller previously proposed for positioning is investigated, and a linear PID feedback controller which is capable of not only positioning but also trajectory tracking is obtained.>

To make a robot track a given desired motion trajectory, a learning control scheme is proposed which is based on the repeatability of robot motion. In this scheme the robot obtains a desired motion by repeating trials (test motion). A merit of this control scheme is that the input torque pattern that generates the desired motion can be formed without estimating the physical parameters of robot dynamics. In practice, to allow the robot motion to approach the desired one in each trial, the input torque given to the robot at the present trial is modified only by the velocity signal of the real robot motion at the previous trial and the desired one. The convergence to the desired motion is theoretically proved for a linear time-varying mechanical system, which is an approximate representation of nonlinear robot dynamics in the vicinity of the desired motion. The effectiveness of this control scheme is demonstrated through actual experiments in which a revolute-type manipulator with three degrees of freedom is used, and the desired motion trajectory is given not only in terms of joint-angle coordinates but also in terms of task-oriented coordinates.>