McKibben pneumatic actuator(MPA), which is one of the soft actuators, has various useful features as an actuator for robots. Therefore, many dynamic motions by a robot using MPA as an actuator have been realized. In our previous research, we also have developed a jumping robot using MPA, but it was a simple robot with only a knee joint. In this research, we develop a new legged robot with an ankle joint in addition to a knee joint. Besides, the biarticular muscle is considered to be useful for the proper interlocking of the knee and ankle joints to achieve higher jumps. Therefore, a biarticular muscle equivalent to the gastrocnemius of humans is implemented in the developed robot. Through jumping experiments using the developed legged robot, the effect of the biarticular muscle for jumping is verified.

Recently, home medical care is being promoted, and the development of a machine that can easily exercise the range of motion of the ankle joint at home is desired. In this study, we have developed a home-based ankle rehabilitation machine using rubber-less artificial muscle, which is lightweight, high power, and excellent in flexibility. In controlling the posture of the rehabilitation machine, smooth operation is required in order to exercise the range of motion of the joint effectively. In this paper, we consider the posture control method of the ankle rehabilitation machine using the rubber-less artificial muscle manufactured.

We propose a novel type of a suction cup capable of adhering to uneven surfaces and wire meshes by the viscosity of MR fluid. In the adsorption principle of the fluid, plasticity, negative pressure and wettability of the fluid are involved. In order to achieve high adsorption force, M-MRF which is MR fluid mixing silica particles is proposed. We show that adsorption to air-permeable material is possible using M-MRF. Furthermore, a mechanism to switch between adsorption and desorption by controlling the fluid properties and transforming the shape of the device is also proposed and the effectiveness of the proposed device is examined experimentally.

In this paper, we propose a new pneumatic actuator to realize the transport of the human body with a flexible and thin structure. At the nursing care site, the goal is to spread the sheet on a chair or bed like a sheet and to gently move the human body. Focusing on the flexible movement of skeletal organisms, we have developed an actuator (Wavy-sheet III) using longitudinal waves that creates a partial contraction in the sheet and sends it to generate propulsion. Compared to the conventional sheet type actuator using transverse waves (Wavy-sheet I, II), it can maintain the transfer function even under high load. The thickness when non-pressurized is about 1 mm, when pressed about 10 mm, and the material is urethane.

This paper describes the development of a new valve that performs the intake and exhaust of a closed flow path by self-excited vibration instead of using electricity. Recently, soft fluid powered robots are developed actively, because of its high power density, lightweight and flexibility. But the more complex the robot, the larger the number of valves and supply tubes required. So in this paper, I propose a valve that self-forces the compression and decompression, and show the results of several performance experiments. We propose a valve that can drive multiple actuators, and show how to remotely control the actuator phase using an air signal.

It is widely known that musculoskeletal robots driven using McKibben pneumatic actuators can realize dynamic motion such as walking and jumping by simple controls. However, the design of pressure inputs of these robots is depended on trial and error, and a quantitative design method has not yet been established. In this paper, we studied a toe trajectory control method for a 2 DOF legged robot with four MPAs. We proposed a quantitative design method of input pressures for any periodic leg trajectory, and confirmed the validity of the method via simulations and experiments. From simulations and experimental results, it was confirmed that the toe trajectory mostly converged to a reference trajectory when the antagonistic muscle had strong antagonism. On the other hand, due to the effects of modeling errors, the deviation from the target trajectory increased as the antagonism weakened.

This paper presents a small type body weight support system driven by rubberless artificial muscle. This system adopts a seesaw mechanism, with one side being a saddle and the other side being towed using rubberless artificial muscle. We performed a walking experiment using this system and evaluated the muscle activity of the gastrocnemius and vastus lateralis of the user. As a result, it was suggested that the proposed system could reduce the burden during walking.

We proposed a fluid-driven jamming mechanism that realizes a variable stiffness mechanism by using pressurization. In this mechanism, a rubber tube is enclosed inside a hollow linear body. When this tube is expanded by fluid pressure, the linear body can be held in any shape. Since the fluid pressure is used, the force can be transmitted even if the deformation is large, and the holding torque can be maximized. When a joint diameter of 17.8 mm and a pressure of 0.2 MPa were applied, a maximum holding torque of 0.21 Nm was obtained. In this paper, the basic principle of the devised fluid jamming mechanism is shown, and the effectiveness of this mechanism is demonstrated by obtaining the holding torque using the prototyped machine.

This paper proposes a new device for measuring the axial length of MPA(McKibben Pneumatic Actuator). The device consists of soft material, hall sensor, and magnet. As the input pressure of MPA increases, the radial length of MPA is more prolonged, and the axial length becomes shorter. By using the relationship, the proposal device measures the axial length of MPA from the radial length. The device is made of soft material and has a donut shape to measure the radial length of MPA. A hall sensor and magnet are mounted outside and inside the donut, respectively. The bigger the device transforms, the closer the distance between the hall sensor and magnet is. First, this paper introduces the recipe of the device. Next, the result of the verification experiments of the effectiveness of the device is described. This experiment was conducted while changing the pressure input and the length of MPA. As a result, it is confirmed that the device has enough performance to estimate the axial length of MPA.

This study proposes a novel inchworm locomotion mechanism for in-pipe robots. Locomotion mechanisms for in-pipe robots must satisfy multiple limitations including compactness, flexibility, high locomotion speed, explosion-proof, etc. In this paper, a novel pneumatic inchworm mechanism is proposed to achieve high locomotion performance under such limitations. A unique duplex chambered structure of the mechanism generates smooth bi-directional locomotion by using pressure changes of two air supply lines. Its advantageous structure also contributes to compactness, lightness, and flexibility of the mechanism. The experimental results showed that the mechanism successfully locomotes through 25-mm horizontal/vertical pipes, achieving 45.5 mm/s of averaged speed in horizontal pipes.

This paper describes about a half rising posture assistive device using pneumatic artificial muscle. This device uses compressibility of air, and it can generate assistive force passively according to the posture of the user. This device reduces an extra load on a disk of the waist by generating assistive force like pushing the chest of the user. In this paper, a basic mechanism of this device and pneumatic artificial muscle which is used in this device are explained, and it is shown the simulation result of assistive force. In addition, muscle activity was measured experimentally when the user who is wearing the device becomes half rising posture. We confirmed experimentally that this device can reduce muscle activity when the user becomes half rising posture.

This paper describes multi-segmentation of artificial muscles integrally molded by a 3D printer and simplification of post-processing. In recent years, with the improvement in performance of 3D printers, it has become possible to integrally mold flexible and hard materials. From this, attention has been focused on the integral molding of actuators that makes effective use of the complex structure of 3D printers. Most of the actuators using 3D printers use air pressure, and in consideration of the material properties, pistons using a bellows structure are common. However, the output of the piston is small because the pressure receiving area is small. Therefore, the pneumatic artificial muscle was integrally molded using the structural advantage. In this paper, we focused on the contraction length of the artificial muscle, and integrally molded the artificial muscle with the increased number of segments. In addition, the design was changed to reduce post-processing after molding. We also discussed the effect of increasing the number of segments on contraction by measuring the contraction of the artificial muscle.

In this study, we have developed a walking rehabilitation support device that attaches to the ankle joint directly to give an assist or load torque during walking using a pneumatic actuator. The effect of the device was verified by measuring the ankle angle and the myoelectricity of the lower leg simultaneously. Torque control is applied in both of assist and loading mode, while an impedance control is implemented in loading mode. We confirmed through some experiments that the assist force can be applied along with the motion of walking and the load resistance also can be arbitrarily adjusted in plantar flexion and dorsiflexion during walking, which show the usefulness of this device.

Existing gripping assist devices in the market is the type of power assist gloves. However, the problem is that the device becomes large and it takes time to mount. In this study, we focus on the pinch motion with an the index finger and thumb, and develop a device that support gripping force triggered by other fingers. The developed device should be compact and easily removable, namely that can be used when assistance is needed and can be easily removed after the operation is completed. A non-powered type, a power type with pneumatic drive and powered type with water pressure. The effectiveness of three developed devices are verified through some experiments.

In our previous studies, it was found that myoelectric potential of biceps was changed irregularly during forearm lifting regardless of the operating conditions. On the other hand, a power assist device that achieves an assist input with a certain tendency has been developed. Although a certain assistive effect by the previous power assistive device is confirmed, this device could not change a tendency of muscle activity which may cause of fatigue of biceps during object lifting motion. Therefore, in this study, we investigate the possibility to control of muscle activity of biceps during object lifting motion by using EMG feedback system.

A snake-type robot using water injection is expected to be applied to fire fighting activities of a disaster. We have developed a Dragon FireFighter (DFF), a flexible snake-like robot using water jets for fire fighting. In this research, we improve the DFF reliability for increase the length of its. In this paper, we developed a metal nozzle module, switched the flow path using a swivel joint, and developed a waterproof case for the motor. Basic experiments were performed using the developed nozzle using the swivel joint. As a result, the reaction force showed a value close to the theoretical value. In addition, it was confirmed that the frequency characteristics did not change due to the pressure change. Using the results of the basic experiment, we proposed a reliable nozzle module.

In order to navigate a mobile robot in dynamic environments such as an attendant robot or self-driving car, path planning using position and velocity of obstacles is ideal. In addition to distance sensors for estimation of obstacle position, tracking between multi measurements is needed for estimation of velocity. 3D motion field generated by tracking points from measurement sensor is called scene flow. We propose stereo scene flow estimation and dynamic object detection to navigate a mobile robot in dynamic environments. The scene flow estimation runs real-time on our target hardware and the dynamic object detection works with some real stereo sequence as input.

This research aims to develop a monocular depth estimation method for robot navigation．In the robot navigation，object detection is important for local path planning．Recently most of autonomous driving cars and autonomous robots use LiDARs or stereo cameras for object detection．However，one of the problems in using a LiDAR for navigation is the cost of the sensor．Comparing to LiDARs，using stereo cameras is much less costly，but it needs to be calibrated properly to get accurate depth data． Therefore in this research we use a monocular camera and adopt a depth estimation using a deep neural network．We evaluate the performance of monocular camera-based depth estimation by comparing with the data from a stereo camera．

Generating realistic data is essential function of Lidar simulation on a mobile robot simulator. However, no opensource mobile robot simulater simulates Lidar intensity. In this paper we propose a new Lidar intensity model based on a Lidar range equation and microfacet models. This model enables the Lidar simulater to simulate reflection and scattering on target objects even with retroreflective materials.

In this paper, a new stereo vision framework that combines the processing results of two cameras and obtains only the three-dimensional (3D) information necessary for each task is used. Under this framework, in a complex environment where there are multiple people, we aim to realize 3D tracking of a specific person by performing person detection using OpenPose. The human region image is divided into upper body and lower body from skeleton information extracted by OpenPose. A specific person is tracked by identifying the person based on the color of the clothes. In this paper, we propose a tracking method and the effectiveness of the tracking method is verified quantitatively by experiments.

The purpose of this study is to make autonomous mobile robots act along human flow in a crowded environment such as stations and downtown areas. As a pre-stage, we propose a method to detect and tracking pedestrians using combining LiDAR point cloud and camera image. In the proposed method, pedestrians are first recognized using Darknet YOLO v2 from the camera image. Then projecting the point cloud data acquired by LiDAR on the camera image, extracting the point cloud of pedestrians only. Further, it is converted into an image overlooking the acquired point cloud, it performs tracking using a particle filter. In this paper, we describe the algorithm of the proposed method and results of implementation by autonomous mobile robot.

In this paper, we propose a trajectory prediction method in an environment where pedestrians are surrounded by obstacles. When an autonomous mobile robot runs around a pedestrian, it is necessary to avoid the pedestrian safely. For that purpose, it is important to predict the trajectory of the pedestrian and take the avoidance action an earlier stage. In this method, the LSTM encoder-decoder model is used to predict the complex behavior of pedestrians. The input of the model is the obstacle information around pedestrian obtained by range-finder sensor and the past position of the pedestrian. Through experiments in a simulation environment, we show that the proposed model can predict the behavior of pedestrians affected by surrounding obstacles.

With the development of technology, opportunities for humans and autonomous mobile robots to coexist in various forms are increasing regardless of the fields. Under these circumstances, needs to improve the comfort of coexisting with autonomous mobile robots arise. However, conventional autonomous navigation that do not recognize humans include elements that are uncomfortable for humans. In this study, I propose an autonomous navigation that recognizes human position and attention to remove such elements. Furthermore, we experimentally evaluated whether the proposed method improved comfort compared to a conventional method. The result of these experiments, proved that the proposed method is more comfortable for humans.

Autonomous robots which move in dynamic environment are required to move to destinations not to hinder the progress of surrounding pedestrians. Previous papers proposed collision avoidance methods based on the position and velocity of pedestrians in the vicinity. However, in a crowded environment, the movement of a robot may affect other pedestrians, or the movement of a pedestrian facing the robot may affect other pedestrians. In addition, by the robot following pedestrians heading in the same direction as the robot, the robot and the surrounding pedestrians can move smoothly. Therefore, in this paper, we propose a harmonious motion strategy that considers the effect of the robot on surrounding pedestrians and the effect of those pedestrians on other pedestrians. The robot determines the avoidance or following action based on the traveling direction of the surrounding pedestrians.

There have been proposed many mobile robots that can run rough terrain. The small mobile robot proposed by us can be applied to various applications, and it is examined to be used for inspection work in remote places in future. It is important to accurately obtain the information around the mobile robot for the operation and autonomous movement in the remote place. However, it is difficult to accurately catch it by the camera, because the inspection object is in the narrow place and there is a difference in height. This paper describes the configuration and concept of the field of view extensibility by mounting a robot arm equipped with a camera on a mobile robot.

This paper describes a study on navigation of an autonomous mobile robot based on metric route information generated from aerial photographs and traversability recognition. Many studies on navigation assume accurate localization. However, navigation using only aerial photographs etc. as prior information causes uncertainty in position estimation. In this study, we take up the problem of traveling on unintended route in the situation where the position estimation of the autonomous mobile robot has uncertainty, and aim at improvement by the navigation system. The proposed method consists of the traversability recognition of the promenade environment, the road boundary detection, and the navigation according to rule of the road.

In this paper, we proposed an algorithm for identifying the elastic property of an elastic rod along its body trunk. The proposed algorithm estimates the stiffness matrix for bending and torsion along the backbone curve of an elastic rod based on the measured information on its shape and the forces and torques applied at its base by making full use of the discretized Kirchhoff elastic rod model and using Newton method. We propose a system for identifying the elastic property as well. Experimental result using an elastic strip with a known elastic property shows that the proposed method can estimate the elastic property of the rod with 0.14 ％ accuracy.

In this paper, we proposed an Au/Pt double layer electrode which makes the soft PEFC durable. We also proposed an expanding internal chamber that can increase the expansion rate and reduce the excessive pressure rise inside the PEFC tube. The resistance of the PEFC tube using the Au/Pt double layer electrodes did not show a notable increase after being bent eight times. The newly developed an air-hose-free thin McKibben muscles using the PEFC tube that has Au/Pt double layer electrodes achieved a maximum contraction rate of 20% and an average contraction time of 31 s. The performance of the air-hose-free thin McKibben muscle such as the contraction ratio was dramatically improved, about four times as compared with the previous muscle.

Rehabilitation gloves have recently attracted considerable attention since the number of patients increases and the number of medical personnel decrease in declining birth rates and rapidly aging society. Especially in the rehabilitation of spasticity, it is important that a large extension torque is generated by rehabilitation gloves in order to sufficiently extend fingers. Therefore, we have developed a soft and lightweight rehabilitation glove with thin McKibben muscles with sufficient extension force. In this paper, we modeled and measured the extension torque of the glove we have proposed. As a result, we achieved to generate several times the torque of a conventional rehabilitation glove.

In this paper, we proposed that the motion analysis system with 6D-MARKER of a snap motor. Many mobile robots are difficult to generate impulse force at a quick time because DC motor generates high power using many gears at low speed. The snap motor is the simple impulse force generate mechanism using the snap-through buckling of an elastic strip and motors. The snap motor can generate quick locomotions for jumping, swimming and so on. We made a motion analysis system used for the snap motor design. The proposed system was composed of a snap motor, the 6D-MAKER (motion tracking system) and a stabilized power supply. We measure the position and the attitude of an elastic strip and estimate the speed.

Previous research on reinforcement learning for continuum robot arms have been dealt with a relatively small number of active degrees of freedom and made experiments of simple tasks such as reaching. We aimed to learn to throw a ball by reinforcement learning in a pneumatically-controlled continuum robot arm that has nine actuators. We adopt Cost-regularized Kernel Regression (CrKR) which uses dynamic movement primitives (DMPs) which is one of the movement primitives. In the simulation, the continuum arm was able to learn how to throw a ball forward. We made the same experiment for our real continuum robot arm and found that the learning progressed in the experiment.

In the prior technology, many soft robot hands used in the food field open and close the hand by pressurizing air. As a drawback of the pressurization method, the film expands at the time of pressurization, and the film may be broken at excessive pressurization. Therefore, this paper uses the decompression method. In the decompression method, since the film acts in the direction of shrinking, the load on the film is small and a larger pressure can be applied. This can be expected to improve gripping force. Even if broken, it works in the direction of suction, so it is hygienic. In addition, since the soft robot hand has a jamming function, it is possible to manufacture a rigid hand while maintaining the flexibility of the soft hand.This is expected to stabilize the grip.

Blood clot forms in a lower limb due to a prolonged period of inactivity. It can be prevented by improving the deep vein circulation through ankle exercises. However, existing automatic exercise assisting devices are bulky. In this study, we designed a novel wearable deep vein thrombosis prevention device using thin McKibben muscles. The device has no rigid structure; it is composed almost entirely of soft materials, is extremely lightweight, and occupies little space around the leg. Ankle exercises, including dorsiexion-plantarexion, inversion-eversion, and abduction-adduction, can be performed in all directions using the proposed device. From the preliminary clinical evaluation, the popliteal vein velocity in the lower limbs, measured using an ultrasonic ow meter, increased after the device assisted ankle exercises. The results indicate that the proposed device can effectively improve blood circulation.

This paper presents dynamic modeling of linear soft robots based on mixed coordinates. We focus on bendable yet inextensible linear soft robots acting on two-dimensional space. Modeling of such linear robots has been based on differential geometry coordinates, but it suffers from complex calculation in kinetic energy formulation. This paper combines differential geometry coordinates and Cartesian coordinates to reduce complex calculation. We show the formulation based on the mixed coordinates with simulation results.

Soft actuators have advantages, including large deformation, safety and adaptability to the environment, and shock absorbance. However, they are weak against sharp objects due to their soft bodies. This paper proposes a protective skin mechanism with an exhaustive arrangement of tiny rigid bodies. Small pieces were sewed on an elastic sheet by Kevlar strings. We conducted some measurements of puncture resistance, elasticity, and descaling resistance of scale mechanisms. The results showed that the proposed scale mechanism had only a 150% larger elasticity than a simple silicone sheet. Also, approximately 15 N was required for descaling, which is seven times large force compared to the glued scale. There was no puncture even when pricked with a needle.

Edible materials have been utilized for soft robots and made them biodegradable and biocompatible. So far, the authors have devised and realized not only soft edible robotic elements but also hard edible materials essential to tasks such as insertion into a narrow environment. However, hard materials tend to require casting mold and thus have infeasible shapes. In this research, In this research, we have devised and devised an edible flexible rigid switching mechanism capable of switching the rigidity state, which was flexible and stiff, respectively, in the research and development of the edible robot we are working on, Characteristics were confirmed by actual experiments.

In this paper, a soft pump which carries fluid in a pulsated fashion is proposed. The proposed soft pump utilizes a snap motor which generates a snap-through buckling of an elastic strip for alternatively shrinking and expanding a soft bag with a pair of one-directional inflow and outflow ports. Due to impulsive motion of the snap motor, the pump can generate a pulsed flow of fluid. The experimental results using a prototype of the proposed soft pump shows that it can generate a very impulsive water flow repeatedly. While the efficiency of this first prototype is not so high thus far, this mechanically simple soft pump has a potential to be used as a propulsion mechanism for a compact underwater robot which swims like an octopus.

Soft robots have attracted much attention in recent years owing to their high adaptability. Long articulated soft robots enable diverse operations, and tip-extending robots that navigate their environment through growth are highly effective in robotic search applications. Since the robot membrane extends from the tip, these robots can lengthen without any friction from the environment. However, the flexibility of the membrane inhibits tip retraction and causes buckling. To resolve this issue, two methods have been proposed; increase the pressure of the internal fluid to enhance the rigidity and mount an actuator at the tip. The disadvantage of the former is that the pressure is limited by the membrane strength, while the second method leads to the robot complexity. In this paper, we present a tip-retraction mechanism without buckling that takes advantage of the friction from the external environment.

The exoskeleton type soft actuator using the 3D printer and the existing flexible tube can change the operation characteristics depending on the skeleton shape, and the airtightness is also guaranteed. Therefore, it is useful as a soft actuator. The exoskeleton type soft actuator can solve the delamination problem by arranging the operation direction and the stacking direction of the FDM type 3D printer perpendicularly. However, in the conventional FDM type 3D printer, there is a limitation on a shape that can be manufactured due to lamination in only one direction. In this paper, we have developed a slicer and a printer that enable stacking in a fan shape. By stacking in a fan shape, it is possible to manufacture an actuator useful for a joint on a cylinder.

Many existing actuators have problems such as high rigidity and low impact resistance. The "Elastic Short Motor", in which the rotor and stator of the motor are connected by an elastic body, has the simplicity of mainly consisting of two parts, but shows various pattern deformations depending on the shape and the connecting direction of the elastic body. The "Elastic Short Motor" can generate an instantaneous force by snap-through buckling, and the impulse is estimated to be 3.05∙10^(-2) [kg∙m/s]. Although this value is smaller than other instantaneous force generation mechanisms, it also has properties such as energy conservation like the Series-Elastic Actuator.

This paper focuses on developing the Fin Ray Effect® based soft gripper, which changes the internal structure and controls the grasping characteristics. The proposed design has movable joints with magnet sheets, and the joint positions are controlled with wire-driven motor system. As results, the proposed soft gripper demonstrated controlling the internal structure. Moreover, the grasping experiments indicated that three different internal structures, which characterizes different grasping conditions.

Soft robots have a problem that they easily punctured or torn by contact with a sharp edge due to their soft bodies made of weak materials. To solve this fundamental issue, we proposed a robotic blood vessel mechanism that enables the robot to heal their bodies themselves. This mechanism realizes the self-healing function of soft robots like blood closing and healing wounds. We have created some test pieces consists of elastomer and inner blood that reacts by two-liquid mixing. In addition, we constructed a twisted helical type vascular mechanism, which is constructed by the gel produced by two-liquid mixing. In this paper, we demonstrate the concept of the robotic blood vessel mechanism and show the results of the basic experiments.

In this research, we propose a scanning haptic sensor that can be used for surface inspection. Minute irregularities on a wide surface must be found in surface inspection. Therefore, we propose a planar scanning haptic sensor with a wide inspection width when user trace once. This sensor has two long strain gauges arranged in parallel. It is easy to detect the signal due to minute irregularities by superimposing the signals of two strain gauges. As prototype, the haptic sensor with an inspection width of 60 mm is fabricated. In this paper, the minute irregularity is defined as a convex of Gaussian-shape that has height of 50 μm and wide of 3mm. We show by experiments that the proposed sensor can detect minute convex.

We are developing clustered convex type telescopic manipulators. Compared with the conventional linier actuators, our convex type manipulator can make long reach and are very light. They are also compact when they are reeled up. Even for these advantages, our previous model type-J, was limited on the power about 100N and stability: this model was our first simple roller-tape driving type and there were some remaining problems which could not be solved. In this report, we will describe the new manipulator type-J2, which employs modified roller driving for the tape, to improve the total performance. The power is about 200-300N and the maintenance is easy.

In this study, we developed a flexible deformable moving body with a two-dimensional shape and aimed at realizing omnidirectional movement. As a previous study, we developed a planar flexible deformation mobile body combining multiple bending modules. It moves by traveling wave which generated by pressurizing the air to the bending module periodically. However, problems such as air leakage from the bending module that constitutes the moving object and the influence of the moving operation due to the tube that is the flow path were also confirmed. To reduce the number of externally connected tubes, we will develop a new bending module that can include multiple channels inside. In this paper, we report the development of a bending module with multiple flow path inside and the results of research on a one-dimensional mobile object composed of it.

Jamming arms have a high degree of freedom, which is one of the main features of soft robots, and stiffness, which is one of the useful features of robot arms. Another useful feature of robot arms is a stretchable structure. However, soft arms with a highly stretchable structure are mainly inflatable arms, which has low stiffness. The goal of this research is to develop a stretchable jamming arm that has a high degree of freedom, stiffness, and stretchable ability. In this paper, we considered basic requirements for a stretchable jamming arm, adopted a grain jamming arm with a unique structure composed of skin and inner grain separately, and made a prototype of skin based on the consideration. The skin of the prototype performed nearly twenty times of stretchable ability.

Power soft robot is considered as one of the potential collaborative robotics for heavy tasks because its compliant structure makes it suitable for human-robot interaction. This study aims to show the high level of safety of a 1-DOF manipulator joint driven by hydraulic muscles in an antagonistic arrangement and its potential to realize a multi-DOF power soft robot. When stored elastic energy is suddenly released, the intrinsic safety of hydraulic muscle actuators with incompressible liquid medium is experimentally proven to be higher than pneumatic muscle actuators. As a result, hydraulic muscle actuator is considered as a better candidate actuator for driving a power soft robot.

In recent years, synthetic fiber ropes attracted much attention because it can increase the load capacity and reduce the size and the weight of the tendon drive mechanism. However, synthetic fiber ropes have a problem of large plastic elongation. That the amount of plastic deformation of the synthetic fiber ropes is very small in the range of loads smaller than the load applied in the past.“Pre-stretching”is performed to remove plastic elongation by applying a load in advance. However, the drawing equipment required for pre-stretching the ropes is very large and unsuitable for research purposes. Therefor, we have developed a stretching device that can pre-stretch even in narrow spaces. It was confirmed that the pre-stretching was done by actually pre-stretching with the developed drawing device and testing the prepared material.

In recent years, it becomes popular to use robots that clean houses. In the future, it is expected that various robots having arm are used at home. But the robot arms used in factories are large and heavy, so it is difficult to use them at home. In this study, an expansion and contraction mechanism, which has a capability of portability and easiness of put-away, has been proposed and the robot arm has been developed using it. The previous robot arm was manually expanded and contracted. In this paper, we improved the robot arm so that can automatically expand and contract and have a wide movable range.

In recent years, in response to the decrease in the working population, especially the number of skilled workers, as a solution to this, smart factories have been introduced at production sites and the number of robots introduced has increased. Smart factories are not progressing. In order to solve this problem, we propose a robot that does not use a cost-reducing reducer that is expensive among the elements of the robot and that can be easily taught by direct teaching. In this manuscript, the design method of the delta type robot proposed up to now and the improved wire feeding part will be explained.