Parkinson’s disease causes various gait disturbances due to dopamine deficiency in the basal ganglia, significantly reducing patients’ ability to perform activities of daily living and diminishing their quality of life. Previous studies have demonstrated that the wearable cyborg Hybrid Assistive Limb (HAL) trunk unit, assisted lateral movement during walking by providing lateral sway, thereby improving gait disturbances. We have developed a hybrid control method for HAL that assists appropriately based on the wearer’s stride time stability and intentional stride time changes, using their biometric data, thus expanding on existing control method. However, HAL controlled by the hybrid control method has not yet been applied to patients, and it is necessary to verify the feasibility of voluntary control that assists in synchronization with the patient’s lateral movement based on motion intention estimated from biometric information. Additionally, the current power transmission link to HAL trunk lacks a rotation mechanism, preventing accommodation of natural trunk rotation during walking and thereby restricting this motion. To address this limitation, we developed a HAL that adapts to the wearer’s trunk rotation during walking. We confirmed through gait experiments on an able-bodied participant that this new mechanism did not inhibit trunk rotation during walking. Furthermore, we confirmed that the developed HAL with voluntary control could assist lateral movement in synchronization with a patient with Parkinson’s disease through the gait experiment.

Locomotive syndrome is a condition in which motor function during walking deteriorates due to motor unit disorders, increasing the risk of requiring nursing care. This is a serious problem that reduces the quality of life. In this study, we focused on cybernics treatment, a method in which a wearable cyborg facilitates voluntary movement in paralyzed body parts, thereby providing sensory feedback from the paralyzed periphery to induce neural plasticity. Facilitating movements that activate the muscles along the spiral line—a myofascial chain that spirals across the body—is expected to improve coordination between the lower limbs and the trunk in patients with impaired locomotion, thereby enhancing their walking function. This study aimed to develop a method for activating the muscle groups along the spiral line responsible for coordinated hip-trunk motion using a wearable cyborg hybrid assistive limbs (HAL) through cybernics treatment, and to confirm the feasibility of this method in assisting twisting movements through coordinated hip-trunk motion in a fundamental experiment. The system consists of a trunk twisting unit and a hip flexion unit. By mechanically linking these components and implementing a control system that synchronizes trunk twisting and hip flexion based on the wearer’s intended movement, the system enables coordinated hip-trunk motion assistance. We conducted a fundamental experiment on an able-bodied adult male. Our results confirmed the presence of assistive torque during trunk twisting and the synchronization between trunk rotation and lateral bending, thereby confirming the feasibility of assisting hip flexion and trunk twisting in accordance with the wearer’s movement intention.

Cybernics treatment using the Hybrid Assistive Limb (HAL) can improve gait abilities. Contrary to conventional evaluations that compare 2-min walk distances without wearing HAL between pre- and post-intervention, we assess gait data measured by HAL during gait assistance, which enables the observation of gait changes that accompany the intervention. To establish this novel evaluation approach, it is essential to examine the relationship between changes in gait ability without HAL and gait patterns during HAL-assisted walking. Focusing on one individual with a neuromuscular disease, this study analyzed and evaluated the relationship between changes in the 2-min walk distance and changes in gait patterns during HAL-assisted walking. Principal component analysis (PCA) was employed to characterize the changes in gait patterns during HAL wear, followed by the creation of an individual model that predicts the rate of change in the 2-min walk distance without HAL from those features using the eXtreme Gradient Boosting (XGBoost). The model was then interpreted using SHapley Additive exPlanations (SHAP) to analyze the contributions of each feature to the prediction. Analysis of nine HAL-assisted walking trials and corresponding 2-min walk distance measurements revealed that changes in the 2-min walk distance were associated with alterations in specific gait patterns during HAL-assisted walking: knee joint angles, knee joint torques generated by HAL, and trunk pitch angles representing anterior-posterior trunk tilting. These findings clarified important features related to changes in gait ability within longitudinal gait pattern changes during cybernics treatment and demonstrated the utility of our analysis method and HAL-measured data for evaluating individual gait pattern changes.

Freezing of Gait (FOG) is one of the typical parkinsonian gait disturbances of progressive neurological disorders such as Parkinson's disease (PD) and progressive supranuclear palsy. Previous studies showed that the wearable cyborg HAL trunk-unit improved their gait disturbances through only an autonomous sway control that provided lateral swing with constant frequency. To promote the improvement by establishing interactive Bio-Feedback loop, it is necessary to realize synchronization between the wearer's intention and gait states for stabilizing gait, i.e. reducing variation of gait cycle. In this study, we developed a voluntary control method that responded to an intentional change in gait cycle and a method that switched 2 kinds of controls including the voluntary control method and the autonomous sway control method for stabilizing gait. The voluntary swing control method synchronized the HAL's force with wearer's gait in a stable gait state less prone to FOG. The autonomous sway control method provided lateral swing with constant frequency as feedback to the patients to achieve a gait state less prone to FOG. These controls were switched according to the gait stability calculated by wearer's gait cycle. Through the gait experiments with an abled-body participant, we confirmed that the HAL's lateral cyclic sway synchronized with the participant's gait and each control switched based on a gait stability. These results showed that the proposed methods had the feasibility of stabilizing gait.

Patients with hand paralysis have motor and sensory dysfunction in their hands. They can neither flex/extend their fingers nor feel gripping force or finger posture. To allow patients with hand paralysis to perform gripping motions, it is necessary to assist finger motion and provide sensory information about the gripping force and posture at the residual sensory area. We previously developed a wearable finger-motion assist system that assists finger motion and measures the gripping force necessary to present sensory information. Therefore, the purpose of this study is to propose and develop a wearable system that provides multiple types of sensory information based on gripping force and finger posture estimated by finger-motion assist system, and to confirm the basic performance of the system through experiments. We developed a band-type sensory feedback system that tightens the forearm in response to gripping force using tendon-driven wires. A vibrator is installed in it, and the intensity of vibration changes in response to finger posture. As basic experiments, identification of gripping force and finger posture using band squeezing and vibration was tested. The results showed the overall identification rate for the combination of gripping force and finger posture that were both correct was 75.6%. The identification rates for gripping force and finger posture when stimulations were presented simultaneously were 93.3 % and 80.9 %, respectively. In conclusion, we confirmed the feasibility of this system that presents multiple types of sensory information.

Patients with knee arthropathy suffer from pain owing to damage at the knee joint cartilage and experience difficulty in walking as the condition worsens. As one of treatments of knee arthropathy, an implantation of cellular cartilage has been performed. After implantation of cellular cartilage, the knee joint should avoid overweight to prevent cell damage form body weight and apply appropriate mechanical stress to articular cartilage to promote cartilage regeneration. Therefore, a weight support system which can adjust the amount of weight in everyday use is required. This study aims to develop an in-clothes body weight support system to protect the knee joint during daily walking to support knee arthropathy treatment and recovery in regenerative medicine, and confirm basic performance of the system through experiments with able-bodied person. The system had frame structure with two-node link from the groin to the foot with the user's knee joint as the center of rotation. It consisted of load-bearing seat, height adjustment mechanism, thigh cuff, mechanical knee joint, and, ground reaction force sensor shoes, and was designed to be sufficiently thin to fit inside clothes. In the 10-Meter walking test, the system could adjust the amount of supporting weight and support two-thirds, one-half, and one-third of total body weight of an able-bodied participant during walking. In conclusion, we confirmed that the developed system had basic performance of weight support.

Decreasing birth rates and an aging population in society often cause labor shortages in the manufacturing industry, making the development of methods to improve productivity based on limited human resources imperative. One way to achieve this is by enhancing product quality via the reduction of product losses and the consequent need for reassembly due to human error. Human error during assembly can arise from specific actions via the arms and fingers. We assumed that these errors can be captured based on the information on human skeletal models. The purpose of this study is to propose and develop a system that acquires information about assembly procedures by using human skeletal models, including the fingers of workers, and notifies them of skipped procedures and errors in part types. Further, we confirm the basic capability of the proposed system via experiments based on simulated assembly work. The proposed system monitors workers' motion based on color and depth images captured by a single RGB-D camera. In addition, we developed a function to detect the process and to point out errors with audible and visual feedbacks when errors were made by workers in assembly processes. In the experiment using simulated assembly work, the proposed system exhibited an accuracy rate of 98.3% with respect to acquisition of assembly processes. In the case of an error in an assembly process, the system was able to point out the error correctly and provide feedback to the worker before he had finished picking up the wrong part. In conclusion, we confirmed that the developed system exhibited the basic capability to acquire work procedures and efficiently point out errors in real time.

Care recipients use a wheelchair or an electric wheelchair because they have difficulty in moving. To support the independent movement of Care recipients indoors, it is necessary that an electric wheelchair performs autonomous movement in narrow spaces and approaches a target whose position frequently changes, such as a chair and observe the physical condition such as the heart function of the care recipients. The purpose of this research is to develop cybernic intelligent mobility system as an autonomous mobile robot wheelchair that has an object recognition function, an autonomous movement function, and a approaching function to correspond the position change of the target object and a vital sensing function to check(detect and manage) the physiological condition of care recipients and to confirm the effectiveness of this mobility system through the basic experiment. The mobility system has a function to find and approach a target such as a desk or chair in a room based on the calculated relative position, even if the initial position of the target is changed. Moreover, a vital sensor was equipped to constantly perform physiological management. To confirm the basic ability of the developed mobility system, we carried out an experiment conducted in a simple living environment assuming a real environment where electric wheelchairs are used. The developed mobility system found the target that was placed in a different position from the initial position in the environmental map. And the mobility system moved autonomously into the area near the target avoiding objects such as walls or furniture. Furthermore, we confirmed basic function to operate the developed mobility system depending on physiological condition. We developed intelligent mobility system that has an object recognition function, an autonomous movement function based on relative position estimation, and a function to find and approach a target. And through the basic experiment, we confirmed the effectiveness of this mobility system.

Patients with hand paralysis often suffer from motor and sensory dysfunction. In many cases, they can neither use their fingers nor feel a gripping force. Thus, they find it difficult to perform certain tasks, such as lifting small objects. It is important for patients with hand paralysis to receive assisted flexion and extension motions for their fingers and gripping-force feedback by leveraging body parts or organs where sensation remains. The purposes of this study are to develop a wearable cybernic finger to assist the gripping force needed to handle objects and to provide gripping-force feedback to patients; the feasibility of the device is confirmed through basic experiments. For the basic experiments, the device was evaluated by a model hand and able-bodied persons. The system successfully assisted the model hand’s fingers to grip, lift, hold, put down, and release cylindrical objects representative of necessary daily objects. Furthermore, six states of gripping force were identified with an average identification rate of 95.6% using vibration feedback. To summarize, a wearable hand-assist unit with sensory feedback has proven to be feasible.

Photoacoustic (PA) or optoacoustic imaging can visualize tissue-optical absorbers, especially hemoglobin, with optical contrast and ultrasound (US)-like resolution and imaging depth. Since both PA and US imaging involves US detection, it is straightforward to develop dual-mode imaging systems with unprecedented functional and structural imaging capabilities. Researchers have already demonstrated the potential of utilizing this complementary contrast for several animal imaging experiments and early clinical pilot studies. PA imaging conventionally uses slow, bulky and high-priced lasers as excitation sources. Use of these high-power pulsed lasers is hindering the clinical translation process of this imaging modality with tremendous potential. Advances in solid-state device technology have recently resulted in the development of a new class of high-power light emitting diodes (LEDs) that can be used as fast, robust and affordable pulsed excitation sources for PA imaging. In this paper, we review multiple LED-based PA/US imaging implementations (commercial and lab-made systems) and demonstrate its functional, molecular and structural imaging capabilities using several clinical/preclinical imaging examples. Specific focus will be given to 2D and 3D superficial vasculature and oxygen saturation imaging in multiple in vivo clinical and preclinical studies.

This paper describes a magnetic resonance imaging (MRI) compatibility assessment of our lower-extremity motion simulator called LoMS which provides gait-like motion for a wearer within his/her lying posture during functional MRI (fMRI) imaging. We confirmed that the existence and the movement of LoMS do not decrease the fMRI image quality when the distance between LoMS and the head coil of MRI is practical distance (more than 400 mm). We also confirmed that LoMS can operate properly in MRI environment, which is to measure own joint angle avoiding the noise from fMRI imaging. Then we show that the brain activity related to gait motion can be imaged during gait-like motion.

Myoelectrical signals have many applications in medical, sports, wearable robotics and computing fields. Wet electrodes are widely used to acquire these signals. In contrast, dry contact electrodes and noncontact capacitive coupling electrodes have been developed. However, their use has several limitations. In this research, we developed a hybrid electrode that is capable of both capacitive and resistive recordings by optimizing the sensor input impedance value using a new electrode noise model that contained noise sources. We extend this design so that noise originated during real usage, such as motion artifacts and noise from electric motors is also measured and removed from the sensor output. In experiments, noise analysis and experiments were performed by measuring myoelectrical signals from both upper and lower limbs in realistic situations, including weight lifting, robot arm control, and walking on a treadmill. As the results, we verified that our electrodes were capable of bioelectrical measurements at noise levels comparable to wet electrodes in realistic situations and with high correlation coefficients between both types of sensors.

The safety of wearable walking assistant robots (W2ARs) is expected to be guaranteed as they are spreading. From our experience of the risk assessments on Robot Suit HAL, we assume that the mechanical angle stoppers and appropriate assembling are inherent safety measures for the W2ARs. These measures prevent the hazardous situations: excess assistance and collision with floor or wall. In this paper, we develop the testing machines to prove their strength. They have the weights imitating a leg or a whole body and simulate cyclic impulsive load during walking by exploiting free fall. We hope this paper helps to develop safer W2ARs and to establish safety standards of the W2ARs.

This paper proposes a novel human-computer interface system, with architecture flexible enough to adapt to various types of physical disabilities and able-bodied person, and also capable of connecting to various devices via a wireless protocol, using a tablet PC as a central system with which a user are interacting. For this interface system surface Electromyogram (EMG) of various muscles which a user still has control, are used as input signals for controlling a tablet. The aim of this system is to extract additional intention of user while user doing deskwork. The characteristic EMG patterns, which are observed when agonistic and antagonistic muscles contract together, are utilized for extracting operational intention. Since these patterns are rarely observed while user doing deskwork or other actions of daily life, this methodology can discriminate the derivation of EMG whether intentional or not. Then the operability of interface system are evaluated by Fitts' law based test GUI. The experimental results show the validity of our proposed interface system comparering with other alternative interface systems.

Several methods have been investigated and realized for operation of exoskeleton robots for assistance of human gait. These systems perform motion intention estimation using the bioelectrical signals of muscle activation, body gestures and kinesiological information, or a mixed combination in a hybrid system. For motion intention estimation of the lower limb(s), information of the lower limbs is usually utilized. However, human gait is not only the function of the lower limbs, but also coordination between upper and lower limbs, adding to balance and cognitive functions as well. In this study, we investigate on how to utilize the synergies of upper and lower limbs of human walking in exoskeleton robot control by using the cane (walking aid). We analyse the synergies of human gait with cane in healthy subjects by means of Principal Component Analysis (PCA) in order to investigate the usability of cane for robot-assisted motor rehabilitation. We also implement a semi autonomous control for an exoskeleton robot, single leg version of HAL (Hybrid Assistive Limb) suit, based on the cane and body joint synergies.

This paper describes a novel motion simulator for the lower extremities of human in a magnetic resonance imaging (MRI) environment. This motion simulator provides a wearer with physical supports to move their lower extremities or physical constrains as well as floor reactive force on bottoms of their feet during gait-like motion in an MRI room so that brain activities could be simultaneously measured. An MRI is one of the most powerful tools to measure activities in any part of brain but a device attached on a subject is limited because the material used for the device should be nonmagnetic. This paper shows the compatibility of the motion simulator that consists of Mckibben-type pneumatic artificial muscles and nonmagnetic materials. Also this paper shows the performance of the floor-reactive-force generator mounted on the soles of the motion simulator.

This paper is concerned with realization of a new kind of three-legged walking machine. The proposed robot has neither coxa joint nor knee joint, so the three legs are just rigidly fixed to each other. Instead of pursuing elaborated joint mechanism for the coxa, we equipped it with two actuators and a pair of weights, so that it can generate torque about the yaw and the pitch axes. Thus it only rocks as a reaction of the mass driving unit. We show that, by appropriate choices of oscillatory controls (i.e., amplitudes, frequencies and phases of the pitch and the yaw actuators), it is indeed possible to achieve rotation and forwarding locomotion of the tripedal robot. In particular, combination of the frequencies selects the rotation/forwarding, while combination of the phases affects the direction of forwarding. We also remark that the realized motion naturally reflects the symmetry of the body structure.

Much attention has been paid to passive dynamic walking as an approach to investigate the walking of human beings and animals. As for quadrupedal passive dynamic walking, it is confirmed that walking resembles that of animals, and that the gait of the robot changes depending on the structure of robot or the environment. Based on these facts, it is conceivable that quadrupedal passive dynamic walking is related to walking of animals, and the walking principle is inherent in passive dynamic walking. In this research, we approach the walking principle through realization of a walking on the level ground by rational-energy input and investigation how gait changes depending on the input. In this paper, a quadrupedal quasi-passive dynamic walking robot named Duke has been developed by applying passive dynamic walking. This robot has only two rolling actuators, which simply provide with rocking motion and not drive the knee or hip joint directly. We conduct walking experiments by Duke with various inputs, and observe its gaits. As a result of walking experiments, it was verified that walking speed was related to the frequency and the phase difference of rocking motion. In addition, through analysis of a shape of soles, we revealed Duke has a nonholonomic constraint that is comparable to that of “two-wheeled robot” on its sole. As a result of analyses based on the nonlinear control theory, we conclude that the sole shape contributes to the transition of walking speed accompanying with change of a phase difference.

Human is capable of adaptive and supple locomotion in the real world characterized by rapid changes, high uncertainly, and limited availability of information. In order to understand the human walking, this work was motivated by the concept of passive dynamic walking robots, which have no actuation or control system except for gravity are capable of stable, human-looking walking. On the other hand, human can produce a stepping motion not only depend on the legs, but also the rotation of the Center of Mass, arm-swing, the motion of the torso and so on. In this paper, a three dimensional quasi-passive dynamic walking provoked by rocking motion in lateral plane has been investigated. The behavioral analyses with the robot experiments show that this robot can walk on a flat ground and a gait speed is related to the period of lateral rocking.

This paper investigates force accuracy of a human finger in three types of support conditions of an exoskeleton. The exoskeleton augments pinching force of a wearer's index finger in proportion to it based on surface electromyography. Three supporting manners of the pinching force are evaluated by switching a fingertip part of the exoskeleton. One is that the assistive force is applied to the wearer's finger so that the force could be sensible by the wearer. Another case is that the assistive force is directly delivered to a grasping object without a wearer's fingertip. The other is that a part of the force directly affects the object and the rest affects the wearer's finger. Through pilot experiments, transitions of the accuracy through training in these cases are compared each other.

Currently, wearers of full body exoskeletons are hindered in their ability to use their upper body as desired due to the rigid back parts used in these devices. In order to maximize their versatility the design and preliminary testing is shown of an exoskeletal spine mechanism, called "exo-spine", that allows the wearer to move all degrees of freedom of his spine and shoulder girdle. Based on the primary forces to be supported during lifting, identified as gravity forces from loads lifted in front of the wearer, as well as functional degrees of freedom, which is a control strategy used by our central nervous system, this mechanism can be actuated using only one motor to provide the required support. Experiments indicate a substantial, although not problematic amount of friction as well as further requirements for the control of the assisting force. Besides improving exoskeletons its basic structure and design principles may be successfully applied to rehabilitation as well.

This paper proposes an active air mat that improves comfortableness and easiness to wear an exoskeleton on a forearm. They are ones of important factors for the system performance evaluation in addition to physical support functions of the system. The active air mat which is installed in interface parts of the exoskeleton enables a wearer to attach and release the exoskeleton in an easy way and a short time by quickly inflating the air chambers to hold a human arm or deflating them to release. The air mat adaptively fills a gap between a human arm of various sizes and an exoskeleton. In addition, the air mat minimizes constriction of blood flow by changing contacting areas with a human arm from the periphery to the trunk. The exhaust air from the deflating air chamber is used to ventilate around skin surface so that humidity of the skin could keep low by the ventilation. The active air mat is evaluated through some experiments from the viewpoint of pressure distribution, blood flow, wearing time, releasing time, body-holding rigidity, and temperature and humidity of a human skin in resting and working states.

Biped walking improves the circulation of blood as well as bone density of the lower limbs, thereby enhancing the quality of life (QOL). It is significant not only to healthy people but also to physically challenged persons such as complete spinal cord injury (SCI) patients. The purpose of this paper is to propose an estimation algorithm that infers the intention related to the forward leg-swing in order to support the gait for complete SCI patients wearing an exoskeleton system called a Hybrid Assistive Limb (HAL), and to verify the effectiveness of the proposed algorithm through a clinical trial. The proposed algorithm infers the patient's intention in synchronization with the deviation of the center of the ground reaction force (CoGRF) that is observed immediately before a person starts walking. The patient conveys this intention by inducing the deviation of the CoGRF, using crutches or handrails with both of his/her arms. In the clinical trial, we confirmed that the algorithm inferred the patient's intention to swing the leg forward, and achieved a smooth gait in synchronization with it. As a result, the gait speed and cadence of the SCI patient with HAL during the 10-meter walking test increased to 6.67 [m/min] and 20 [steps/min], respectively after several trials.

We have developed a "red blood cell (RBC)-photon simulator" to reveal optical propagation in prethrombus blood for various levels of RBC density and aggregation. The simulator investigates optical propagation in the prethrombus blood and will be applied to detect it noninvasively for thrombosis prevention in an earlier stage. In our simulator, Lambert-Beer's law is employed to simulate the absorption of RBCs with hemoglobin, while the Monte Carlo method is applied to simulate scattering through iterative calculations. One advantage of our simulator is that concentrations and distributions of RBCs can be arbitrarily chosen to exhibit the prethrombus, while conventional models cannot. Using the simulator, we found that various levels of RBC density and aggregation have different effects on the optical propagation of near-infrared response light in blood. The same different effects were acquired in in vitro experiments with 12 bovine blood samples, which were performed to evaluate the simulator. We measured RBC density using the clinical hematocrit index and RBC aggregation using activated whole blood clotting time. The experimental results correspond to the simulator results well. Therefore, we could show that our simulator exhibits the correct optical propagation for prethrombus blood and is applicable for the prethrombus detection using multiple detectors.

This paper reports support effects of an exoskeleton system for activities of a hand and an upper limb of a healthy person. The support system augments human forces of a hand, a wrist joint and an elbow joint based on bioelectric potential of each muscle so that the upper limb could be assisted by the exoskeleton with a certain rate of wearer's force. Actuators in the assistive system are replaced with powerful ones to supports human hand, wrist and elbow activities with larger force and torque. Through experiments it was confirmed that a wearer receives physical support from the system for activities of a hand, a wrist and an elbow joint and then we evaluate rate of assistance by comparing the magnitude of the bioelectric potential between under a supported phase and under an unsupported phase.

This paper proposes the basic technologies in order to develop a wearable hand assistive system for daily life support. Current prosthetics have some problems for wearability. We focus on the load caused by weight of wearable system. The actuator is one of the heaviest parts in wearable systems. Therefore we propose to use the McKibben pneumatic artificial muscle which is lightweight and compact in size. At first, we propose a new method to control air pressure of artificial muscle without pressure sensor in order to reduce the size of system. Second, we investigate variable stiffness of human finger to perform human finger dexterity and simulate it by using our proposed polyarticular tendon drive system.

This paper introduces a cooperative control algorithm that designs a stable biped walk satisfying a wearer's intention relating to his/her walk such as a timings to start and stop walking, walking speed and waking direction. Using this algorithm an exoskeletal walking support system could help a paraplegia patient walking comfortablly. At first, a pair of gloves with several DOFs is developed to convey a wearer's intention to the walking support system. He/she swings both his/her index fingers as to simulate foot motions of his/her walking. The amplitude and period of the swing corresponds to a step length and period of the walk, respectively. Pronation/supination of the wrist joint of his/her right arm corresponds to a walk direction. The cooperative control algorithm based on a cart-table model designs trajectories of each joint for stable walking pattern that satisfies the intention expressed by the wearer's hand motion and then the designed walking pattern is executed in realtime by the walking support system. As the first trial, a small humanoid robot ¿HRP-2m¿ is used for safety as a control target that will be a combination of a wearer and the walking support system in the final situation. Through some experiments we confirm that our proposed algorithm enables the humanoid robot to start and stop stable walk with variable step length in the desired walking direction according to operator's intentions.

This paper introduces an exoskeleton assistive hand that supports human hand and wrist activities by using user’s bioelectric potential to control the exoskeleton movement. The exoskeleton has three active joints for an index finger, three active joints for combination of a middle finger, a ring finger and a little finger and two active joints for a thumb. It also has two passive joints between the index finger part and the combined part of the three fingers. Our proposed poly-articular tendon drive mechanism simulates a mechanical compliance of a human finger so that the exoskeleton could realize comfortable and stable grasping. This paper proposes a new mechanism “dual sensing system” and a new control algorithm “bioelectric potential-based switching control” so that the exoskeleton could synchronize wearer’s hand activities without any force sensor. A tendon-driven mechanism and a dual sensing system enable wearer’s fingers to move freely when they does need power assist but precise position control or force control. A bioelectric potential-based switching control enables the exoskeleton to augment their grasping force only when wearer’s fingers generate a relatively large grasping force. A five-parallel-link mechanism is used to assist wrist activities of a wearer. Through experiments it is confirmed that the exoskeleton does not disturb a wear’s pinch of a small object and that it augments grasping force for a heavy work.

This paper introduces a new wearable handling support system for a person who has trouble in motor capability of his or her upper limb. The support system is used as not only a support system to make his upper limb active in daily life but also a rehabilitation system to reduce manual loads of physical therapists. The system measures three rotation angles of patientpsilas head: pitch, roll and yaw to control three degrees of freedom of the support system; angle of elbow joint, angle of wrist joint and hand close/open, respectively. Hemiplegia patient who has paralysis of one half of the patient body can use both arms cooperatively by wearing the handling support system on the paralysis side. In our experiments, the system helps pouring task from POP bottle to a glass, while an upper limb on paralysis side of a user grasps the POP bottle and the other upper limb grasps the glass.

For assisting human motion, assistive devices working as muscles would be useful. A robot suit HAL (hybrid assistive limb) has been developed as an assistive device for lower limbs. Human can appropriately produce muscle contraction torque and control joint viscoelasticity by muscle effort such as co-contraction. Thus, to implement functions equivalent to human muscles using HAL, it is necessary to control viscoelasticity of HAL as well as to produce torque in accordance with operator's intention. Therefore the purpose of this study is to propose a control method of HAL using biological and motion information. In this method, HAL produces torque corresponding to muscle contraction torque by referring to the myoelectricity that is biological information to control operator's muscles. In addition, the viscoelasticities of HAL are adjusted in proportion to operator's viscoelasticity that is estimated from motion information by using an on-line parameter identification method. To evaluate the effectiveness of the proposed method, the method was applied to a swinging motion of a lower leg. When this method was applied, HAL could work like operator's muscles in the swinging motion, and as a consequence, the muscle activities of the operator were reduced. As a result of this experiment, we confirmed the effectiveness of the proposed method.

This paper proposes an algorithm to estimate human intentions during walking. Not only walk start or stop but walking cycle is considered as the intentions in this paper. The algorithm is embedded into a walking support system, a wearable robot "Robot Suit HAL-3", for paraplegia patients. The estimation of patients' intentions is indispensable for effective and comfortable motion support, but the biological signals such as myoelectricity which is used for the support by HAL-3 cannot be measured properly. The proposed algorithm, therefore, estimates patients' intentions from other channels such as a floor reaction force and a body posture. The effectiveness of this algorithm is investigated through experiments with two types of patients. One has a sensory paralysis on both legs, especially a left leg has severe trouble. The other has troubles in sensory and motor ability on both legs. We show HAL-3 supports patients' walk comfortably, estimating patient intentions.

The purpose of this paper is to suggest new 3-D N-link system movement simulator. Previous presented method used to "Jacobian" matrix for connecting between the workspace coordination and the angle state space for representing the whole N-link system. But this system has difficulties for simulating the whole dynamics of the general N-link system from the side of computational cost and the precision of the calculations. The major reasons depend on the fundamental principle of conversion equivalence between the work space coordination x/sup /spl rarr// and the angle state space q/sup /spl rarr//, this principle assure that we can change the work space coordinate into the angle state space coordinate which is convenient for the calculation, but this conversion have singularity that is, there is no inverse matrix of "J". In addition, there is accumulation of calculation error when it follows from the link root to link end in the case of calculating the multi-joint link movement dynamics, and the problem of treating the external force into the angular space dynamics. In this study, we consider the N-link system as N-particle movement system. And each of the particles is connected by a kind of spring damper model in imitation of a link. Because of this, our proposed method has no coordinate conversion between workspace and angular space in the dynamics simulation. And by the way of introducing some dynamics restriction, we can construct many of joint features such as fixed joint or single revolute joint, spherical link and so on. And we confirmed the 3D N-link system performance through some dynamics simulation such as walking, this simulator shows good accuracy and computational cost performances comparing previous proposed N-link system simulator.

It is clear that many creatures including human being and macaque have dealt with multi joint arm movement by taking a special control strategy from many biological experiments. These studies said that when moving the hand between pairs of targets, subjects tended to generate roughly straight hand trajectories with single-peaked, bell-shaped speed profiles. For explaining this experimental result, many of movement control strategies have been proposed so far, though, there are many of the problems for resolving multi-joint arm control. Kawato said that the major problem is collected into "two ill-posed" problems. First is the way of selecting a single trajectory from the possible movement. Second is the control problem for realizing the planned trajectory by deciding each of the joint torque and for reaching the whole actuator system to an objective position. First of all, we argue that what advantage a creature gets by taking a bell-shaped speed profiles control strategy form the viewpoint of simple single particle movement experiments and discuss the way of realizing the control procedure. In this paper, we proposed a new control strategy which the speed profiles of the end-effector is modified as the bell-shaped form actively. We regard the end-effector of a multi joint system as a kind of single particle system in this method. By the subsumption, the control problem of the multi-joint system movement can be simplified. And we will show that the proposed strategy accomplish the control of a two-link arm reaching movement, and after that, realize a standing up movement of human model by constructing four link systems which use the physical parameter of human being.

We have developed the exoskeletal robotics suite HAL (Hybrid Assistive Leg) which is integrated with human and assists suitable power for lower limb of people with gait disorder. This study proposes the method of assist motion and assist torque to realize a power assist corresponding to the operator's intention. In the method of assist motion, we adopted Phase Sequence control which generates a series of assist motions by transiting some simple basic motions called Phase. We used the feedback controller to adjust the assist torque to maintain myoelectricity signals which were generated while performing the power assist walking. The experiment results showed the effective power assist according to operator's intention by using these control, methods.

It is thought that the dynamic motions with lower limbs such as walking can be made more effectively by moving each joint along to its natural frequency. If the natural frequency were able to be modified purposefully, the effective motion could be performed at any angular frequency around joint. We realized that with the exoskeleton-type powered suit, HAL (Hybrid Assistive Leg)-3 we developed for walking aid. In this research, we considered the operator's leg as pendulum model, identified the physical parameters around human's knee joints, tried to adjust the impedance, and applied that to pendular movement of leg. The effectiveness of adjusting the natural frequency in power assist control can be confirmed through the experiments evaluated with myoelectricity.

To develop an effective medical care with the internal medical robot, artificial heart, we proposed an online parameter identification of systemic circulation model using delta-operator, which can calculate the time varying and unmeasured hemodynamics of the internal human body from some measured physiological data: Aop (aortic pressure), AoFlow (aortic flow) and Pflow (pump flow). This method consists of: (1) the systemic circulation model, which was configured with Ca (aortic compliance), Ra (aortic resistance), L (aortic inertia) and Rp (total peripheral resistance); and (2) system identification using delta operator. In the computer simulation, we confirm the effectiveness of the proposed method. In animal experiment with the left ventricular assist system, the physiological parameters were identified as: Ra=0.04 mmHg sec/ml, Ca=0.7 mmHg/ml, L=0.02 mmHg sec sec/ml, Rp=0.3 mmHg sec/ml. By the identified parameters, we estimate the physiological behaviors. This method will bring the new stage of development of the artificial heart.

This paper describes the power assist control for walking aid based on EMG and impedance adjustment with HAL-3 we have developed. Virtual torque derived from EMG is adopted as a basic control method, and the motion assist control as to operator's intention can be realized by this method. And we suggest the impedance adjustment around knee joint for more effective power assist control. Experiments for simple motion and walking motion were performed to verify the proposed approach, with impedance parameters found by RLS (recursive least square) method. The evaluation of assisted motion was done by a calculation based on EMG in nearly proportion to the operator's muscle force. The results showed the amplitudes of EMG were reduced significantly, the operator was able to swing the leg lighter by reducing the inertia around knee, and the strain of knee in foot-grounding could be alleviated by adding the stiffness to joint.

Biological experiment results said that a trajectory of the reaching arm movement is roughly straight path in unconstrained point-to-point movement. Many manipulator control methods have been achieved by defining an energy or potential index such as minimum-jerk or mean squared residual. However, in almost all the cases, the potential or energy like approaches have some disadvantages especially for controlling dynamical system, since the energy index is constructed in order to treat an equilibrium state, not dynamical one. On the other hand, an actual manipulator movement is commonly a dynamical control task. To overcome this problem, we introduce the dynamical properties into a stable index. In this paper, we propose a new type of the motor controlling method of a multi-joint manipulator based on the biological experiment results by defining a new type of stable index. In addition, we show the way to control any kind of multi-joint manipulator by using one simple principle, even under external perturbations like the gravity effect simultaneously.

This paper presents a virtual humanoid robot platform (V-HRP for short) on which we can develop the identical controller for a virtual humanoid robot and its real counterpart. The unification of the controllers for the virtual and real robot has been realized by introducing software adapters for two robots respectively and employing ART-Linux on which real-time processing is available at the user level. Thanks to the unification, the controllers can share softwares with the dynamics simulator of V-HRP, including the parameter parser, kinematics and dynamics computations and the collision detector. This feature can make the development of the controllers more efficient and the developed controllers more reliable.

Generally, the operator's force signal is utilized in the case of the power assist. However, the operator must keep always providing the force in order to obtain the force signal. We have developed an exo-skeleton type power assist system, HAL (Hybrid Assistive Leg) for the walking aid. The purposes of this study are (1) to estimate the operator's action in the initial stage of the motion, (2) to propose the method for realizing the power assist corresponding to the operator's motion. We include the following mechanism into HAL system: (1) the mechanism which estimate the reference input of the bend angle of a joint by using myoelectricity, (2) the mechanism which autonomously generates the motion using phase sequence. For optional steps, experiments were performed to realize the stepping-up motion which reflects the intention of the operator. As the result, HAL started the autonomous motion immediately when the operator intended to start the motion, and stepping-up motion which reflected the intention of the operator was able to be achieved, even if the operator hardly generated the power.

The purpose of this research is to propose a quantitative analysis method to analyze the relationship between the feedforward controller and feedback controller in a motion learning process, and also to assess the effectiveness of this method by experiments. To analyze the mechanisms of the motion control systems, we propose: 1) a control gain identification method, 2) a squared-error separation method, 3) a gradient separation method, and 4) a pattern correlation method. The inverted pendulum is used to evaluate the effectiveness of these methods. Experiments show that the proposed methods are an effective way to separate the motor control system into feedforward control and feedback control. They also disclose the fact that there are two learning processes; one of which mainly performs feedback control and the other mainly feedforward control.