The aim of this study is to develop an effective anchor for soft ground at disaster sites. Inspired by the roots of plants, we have developed the “ Root-Anchor”, which is anchored by extending several thin wires into the soil. In addition, in order to clarify the design method of the anchor, we conducted experimental verification of the effects of the number of wires and the angle of the wire insertion on the anchoring force.

The purpose of this study is to realize a remote-control system to implement autonomous flight control of engine-driven multicopter for forest survey system. A cooperative control of engine-speed and blade pitch angle control is important for movement of the engine-driven multicopter. A laser sensor mounted on the multicopter is used for forest survey. In this paper, we model the multicopter to realize its cooperative control. We confirmed the communication between the proportional transmitter and the flight controller to build a remote-control system. We integrated the two communication systems to operate the motor from the PC. The motor operation was confirmed, and the remote-control system was constructed.

This research aims to develop an autonomous vehicle for measuring soil moisture content in forest.In this research, we investigate a suitable vehicle and sensors for autonomous driving in forest for the purpose. In this direction, we utilize a remote-control vehicle with noncontact sensors for soil moisture measurement. This preparation contributes the development of the vehicle driving autonomously, measuring the soil moisture content with noncontact sensors in forest.

In recent year, research on swarm robots has been actively carried out. Projects are underway to develop swarm robots for use in environments where human intervention is difficult, such as outer space and the deep sea. However, such an environment requires a high degree of autonomy and task efficiency to be maintained. In this paper, we focus on the shepherding task, and propose a model that autonomously adapts to the characteristics of the target group and guides it. The submittion of a guidance algorithm that can maintain autonomy over a long period of time will enable applications such as debris removal in space, resource exploration, and marine debris removal.

In final waste disposal sites, after the waste was buried in the ground, it produces gas in the process of decomposition. Degassing pipes are installed at the final disposal site, and the gas concentration is measured regularly in order to check the progress of purification. For labor saving, it is desired that the autonomous vehicle for environmental survey travels to the gas vent pipe and automatically measures the gas concentration. To investigate rough terrain efficiently, we have developed a prototype utilizing a four-wheeled ATV and proposed a control method named Path-Generating Regulator (PGR). This paper focuses on the detection of the gas vent pipe as a destination of the ATV, proposes an algorithm to detect the gas vent pipe. In addition, we will confirm the detection accuracy of the gas vent pipe.

In this study, we have been investigating a cart equipped with an omni-wheel type moving mechanism for ground-penetrating radar surveys. In order to investigate a power-assist control method for a measurement cart in an outdoor environment, an experimental cart was constructed and the behavior of the cart was measured using load cells and acceleration sensors. As a result, it was found that the measurement cart was exert to cyclic forces when the operator pushed it by hands. Based on these results, the bogie was improved.

In impedance control, impedance parameters can be adjusted to improve the performance of human-robot collaboration. This paper proposes a new method to adapt the damping parameter based on the potential field method widely used in path planning. First, a damping field is generated to modify the damping value based on the current end-effector position. Bayesian optimization is then used to learn the parameters of the constructed damping field. Finally, experiments are conducted with a 2-DOF planar robot arm in a point-to-point collaborative motion, and results show that the proposed method help improve the performance of tasks that require high accuracy.

Carts and carry bags are used to transport goods, and they place a heavy burden on the body, especially when ascending or descending stairs. However, crawlers are not suitable for stair climbing because they cause slippage. The driving unit of this cart has a planetary gear mechanism with three drive wheels on each side, and it can move on a plane by rotating the wheels and climb stairs by rotating the arm part. As a result of the power-assist performance on a plane and stairs, the burden of the user was reduced at both places, however, there was a problem of poor operability for descending stairs. In this study, we propose a new power-assist control method to solve the problem in descending stairs and evaluate the method through experiments.

The demand for collaborative robots is expected to increase in the future due to their high adapt-ability. However, most of collaborative robots available in the market are driven by a combination of electromagnetic motors and reduction gears, and thus the weight of the arm itself tends to become larger in order to increase its payload. This study focuses on the hydraulic robot arm, which has an advantage in the force/weight ratio, as a means of realizing collaborative robots for heavy objects. This paper proposes a robot arm that reduces its tip speed while maintaining the maximum force by limiting the flow rate to each motor of the arm. This paper also confirmed the applicability of hydraulic robot arm to actual collaborative work by conducting a demonstration experiment in which an operator attaches a 20 kg steel plate to the target position using the developed robot arm.

Robot safety is the most essential element when people tried programming robot working alongside themselves. In this work, an approach to the intrinsic safety functions is introduced. Nicebot-5 has six joints, and each joint has been installed with an electrical friction clutch. Clutches have a feature that they could slip when torque exceeds a defined value of duty, which indicates that they can be used as torque limiters for robot safety. By setting minimal torque limits for different kinds of motions, the robot is expected to complete the desired trajectory unhindered. At the same time, it could immediately stop if the robot collides with objects. The experiment process section explains how I calculate the minimum torque limits for all movements by defining simplified dynamic and clutch models, measurement experiments, and data analysis. As a result, Nicebot-5 achieves intrinsic safety with the motion of all orientations.

Recently, wearable robots, which add extra limbs like arms or legs, have been researched for the purpose of human augmentation. However, these wearable approaches, due to their weight and configuration of themselves, affect the inertia of the users and interfere with users walking. In this research, we propose a novel form of a mobile robot that is designed to surround a user non-contactly. This paper provides findings about the design of control performance requirements that we obtained from a walking task experiment. This robot can be prospected to use for human augmentation using the user’s peripheral space.

The representation of avatars in virtual environments (VE) affects co-presence was reported. These reports are limited only to visual interaction. This paper aims to clarify the effect of the avatar of representation on the interaction force in the cooperative task in VE. The interaction force in VE is essential for the construction of teleoperation. The cooperative task with multilateral control was conducted, and the interaction force and questionnaire were analyzed. As a result, the avatar’s representation was not affected by the interaction force. On the other hand, the subjective evaluation was changed by the avatar of representation. Therefore, it was concluded that avatars are not always necessary in a cooperative task.

We often use carts to carry luggage at the shopping center and airport. However, when the weight of the cart is heavy or the cart is surrounded by a lot of obstacles, the burden of the human controlling the cart increases. One of the research for reduction of the burden is the control using attractor that divides the work into human for sensing and propulsion, and machine for turning. In this paper, we propose a design method of vector field by extending dimension regarding velocity. Using the proposed method, it’s confirmed that the cart moves on reference trajectory regardless of the direction of motion and the steering wheel doesn’t rapidly change when we switch the direction of the motion.

Over recent years in particular, the assistance systems which help the human operation and behavior such as driving have grown at an incredible rate. It is undoubtedly believed to have a major impact on the way people work cooperatively with the system. Although machines are considered to often out-achieve humans operationally, they could not assume control or responsibility and attain professions that would once have been out of bounds. Therefore, we are in favour of the theory, the frame problem, and focused on human reliance on the assistance system. We aimed to initially estimate human reliance doing a task with the system.

The conventional remote operations of an excavator have low work efficiency compared with direct operations. To overcome this problem, we have developed a semi-autonomous control system that consists of autonomy (dynamical system with attractor) and human action (by admittance control). Furthermore, we have designed discrete task selection in the semi-autonomous leader–follower system using a nonlinear dynamical system with bifurcated trajectory attractor, however, it was difficult to achieve task trajectory intersection and proximity. In this paper, we design a semi-autonomous control system that avoids the intersection and proximity of trajectories by dimensional extension of the dynamical system and realizes the selection from similar tasks. The proposed system is experimentally verified using a leader–follower excavation robot.

Haptic Feedback Belt is the device which notice the supernumerary arm’s posture which is put on the arm’s user as vibration feedback. In this paper, we examine whether participants can perceive the supernumerary arm’s posture which leaves them with Haptic Feedback Belt on Dual task. Experiment was conducted to compare perception accuracy when they accomplished the recognizing task of supernumerary arm’s posture on Single task situation and Dual task situation. On Single task situation, they accomplished only the recognizing task. Then, on Dual task situation, they accomplished the recognizing task with pouring water into beakers with their own body. The results suggested that they can perceive the posture on Dual task situation as exactly as on Single task situation.

Society 5.0 shows a human-centered society, where everyone can participate in a variety of social activities by supporting humans with AI and robots. However, in human-assisted technologies, human operation input and controller input are entered the control target simultaneously, and the input signals are interfering with each other. Therefore, the controller must recognize the human operation input information before it can operate to obtain the desired response. In this study, database-driven control method to deal with dynamically changing human operations is proposed. The effectiveness of the method is verified by numerical simulation.

In this paper, the authors report on the development of creating a software module that can make arbitrary web page have contactless operation feature with neural network model. To obtain hand coordinate data for training, the authors created a special web application with MediaPipe which is an open-source library provided by Google. Thanks to the application, sufficient number of data can be obtained easily. As a result, the authors confirmed that the model can provide excellent UX. We then experimented to evaluate the performance of a software module used the model and confirmed that how accurately it could work with that module.

It is desirable to use assistive robots to support daily tasks performed by humans. In this study, we conducted a routine patrol to manipulate and transport objects such as garbage disposal. By coordinating multiple robots with different characteristics, it is possible to perform long daily patrols, object manipulation, and transportation, which are challenges in performing such patrols. By using voice for the task handover process between robots, we were able to build a system that is robust against failures due to human intervention, as well as scalable to adapt to new robots by simply starting the task with voice commands.

To perform a certain task smoothly, it is necessary for them to become proficient of a task through the training and practice. If training can be executed in a virtual space, it will be possible to train anytime and anywhere. In the training of tasks, avatars in the virtual space have been used. In the case of the virtual space, it is expected to promptly improve proficiency, reduce the cost and psychological burden comparing conventional physical training. However, the virtual collaborative tasks have not been studied, yet. In the training of collaborative tasks, the use of avatars in the virtual space are needed. It will be possible to train the work even by one operatorif avatars join the training in the case of collaborative tasks. In this paper, a collaborative task between a physical operator and an avatar in the virtual space using HMD is proposed. A physical operator wears a VR glove in MR, and the effectiveness of repeated practice training is evaluated.

In this paper, the authors aim to correct strokes for input of handwriting diagrams. In order to achieve this correction, they propose a method which is in real-time using machine learning. Experiments with this method have shown that machine learning is also effective in stabilization of handwriting.

Rhythm is an important element in communication. For example, it plays a key role in a wide range of fields such as conversation and collaborative work. Therefore, rhythm is an important issue in a society where robots coexist. The purpose of this study is to construct a system that Robot controls so that it is not noticed and synchronizes human rhythm. In addition, in this paper, in order to realize this, we investigated the easiness of human rhythm. The result shows that when a slow tempo stimulus is given while the tempo is maintained, there are patterns in which the maintained tempo becomes faster or slower.

Recently, With the development of science and technology, automatic systems have become waidely used. Automatic systems are being used for operation support systems.Automatic operation support systems are expected to replace human work, reduce the workload, and suppress human errors.However, automatic systems can’t deal with all ploblems. Therefore, cooperative work with human-centered automatic systems is attracting attention. In order to be human-centered, it is important to consider the ”sense of self-centeredness,” and research is being done to evaluate it. However, conventional evaluation methods are difficult to generalize to actual dynamically changing systems such as driving. In this paper, we purpose assessing a consecutive Sense of Agency by measuring operational performance in machine operation.

Due to the stereotype ideas and lifestyle in Japan, unbalanced household working time between males and females has been kept for a long time. In this study, we have proposed the use of biosensing to identify the emotions of infants and communication robots to assist child-care tasks. As preliminary trials, we developed a heart rate measurement system with combining a communication robot and conducted two experiments. The results of the experiment 1 show LF/HF could be an indicator to estimate infant emotions. According to the results of statistical analysis in the experiment 2, LF/HF has an increasing tendency after the cheering message by communication robots, although there are not significant differences.

This paper presents a development of light-emitting chair with an internal light source that emits light in synchronization with individual’s heartbeat and respiration rhythm. We assumed that the visual stimuli can affect interoception because the light emission pattern is based on the vital signs. Moreover, in color psychology, colors in surroundings can affect human emotional states and behavioral responses. The purpose of this study is to discuss novel approach to bring a feeling of relaxation by using a light-emitting chair. Furthermore, we propose a method for monitoring respiration signal from time-series data measured by the pneumatic pressure sensor that inserted into a cushion on the chair.

Some social robots can provide people with well-being. We have developed a flying social robot “ Hubo ”that provides more well-being. Hubo, whose design is simple and characterless, flies around randomly and moves in the direction of the sound. The design and actions can give people room for interpretation. In this paper, we propose its design and verify its performance. The results imply two conclusion: (i) Hubo is able to fly forward and circularly, (ii) Hubo is able to estimate the direction of human voices.

It has been reported that interactions through robots have a positive effects on people. In particular, robots whose exteriors are made of soft materials have been found to create impressions in interactions such as communication with people based on their first impressions. In addition, it has been shown that the use of robots with characterization has a positive impact on such impressions. The robot used in this study has a soft exterior and a characterization based on Professor Donmasu, the imaged character of Toyama Prefectural University. The lack of integrated control of the individual functions of this robot causes problems in terms of operability, safety and prevents the robot from performing interactions that integrate the individual functions. Therefore, the purpose of this study is to develop a control system that integrates individual functions.

In this study, we evaluated the impression of the beckoning motion of the robot. Seven kinds of beckoning motions were generated by different motion times for swing down and swing up phases, respectively. We obtained the emotional differences of subjects among seven motions by semantic differentials method, and the data were analyzed by means of a factor analysis. 21 subjects in their twenties participated in this experiment. As the results, our impressions related to aggressiveness and comfort were affected by the motion parameters. Especially we inferred that total motion time consists of swing down and up phases affected these impressions.

The development of remote robots that reduce the amount of human movement will increase the amount of time humans can engage in intellectual activities. In previous research, human-robot interaction has included voice and motion instruction. For communication between humans and remote robots, it is essential to communicate in a format that is easy for both sides to understand. In this paper, we propose a system that enables users and robots to interact remotely through an Internet chat system using text and images that are easy for humans to understand, and buttons that are easy for robots to understand, in situations where users would normally have to select actions locally.

In order to improve the convenience of communication robot arms compatible with smartphones, it is suitable to reduce their size. Shape memory alloys (SMAs) are excellent actuators for small robots due to their adaptability to miniaturization and large deformations. By applying SMA to the joints of a small robot arm, the movement mechanism of the robot can be made smaller than the power of a motor or the like. In this research, we proposed a small gesture robot arm that can be controlled by a smartphone. We attempted control of the drive angle of a small gesture robot arm driven by SMA wire by serial communication from a smartphone. Evaluation is performed based on how well the required angle input to the robot system matches the actual drive angle of the robot arm.

Annotation work is necessary for natural language processing based on machine learning for robots. For the purpose to realize the automation of this task, we studied some automatic labeling methods for product review evaluation, and proposed two methods using the analysis of adjectives and learning of human ratings. The proposed two methods were evaluated by experiments and their results were compared. Based on these results, we also discussed the usefulness of the proposed methods when applying to other datasets.

When using knee orthoses, the user feels pain by the force due to the misalignment between the user and orthosis. This is mainly due to that the center of rotation (CoR) of the knee does not coincide with the CoR of the orthosis. However, the CoR position of the knee is not constant, and there may be misalignment in the wearing position of the orthosis. In this study, we solved these problems by using a parallelogram linkage. It passively deforms as the knee expands and contracts, thereby absorbing the misalignment of motion. We proposed a design method to minimize the length of the parallelogram linkage that can tolerate the assumed misalignment of the mounting position. Finally, through experiments, we showed that only torque can be transmitted between the orthosis and the leg attachment without generating pain-causing forces.

In this study, a force feedback device that acts directly on finger joints using a tendon-driven mechanism with multiple tendons is developed. Attaching three tendons directly on the finger without links reduces the user’s burden and presents multi-directional force feedback. Furthermore, since it is difficult to model the proposed structure accurately, the moment arm of the tendons for the joint axis was obtained through measurement experiments with the developed device. Finally, we validated the fingertip force while driving the developed device through the experiment. The results showed the possibility of controlling the direction of the force feedback by the proposed method.

Resently, IoT devices have spread thorugh various industries. Therefore Smart Speakers have spread too. Dialogue Spoken System technology is attracting attention. These technologies use speech to estimate the speaker’s intentions. However, the problem is that these system was used to classification estimate intention. So, It was not possible to estimate intent in more detail. In this resarch, We focused on changes in speaker intention and expression. The results, each Speakers representation were suggested.

Overhead work is considered to have a high risk of musculoskeletal disorders, and wearable robots that assist arm movement have been developed, but most of the arm support devices on the market are expensive and have weak assisting power. Against this background, we developed and commercialized an arm support device that uses a gas spring and has as simple a structure as possible. However, due to the characteristics of the gas spring, problems were encountered with difficulty in movement when the device was attached. Therefore, we developed a new arm support device that uses McKibben-type artificial muscles. In addition, a comparison with the gas spring model was conducted. In this paper, we describe the structure of the developed device and the results of evaluation..

Typical sports support methods have a possibility that people will acquire a form which puts a burden on one part of the body or a form which cannot exert sufficient power and a problem that an instructor who gives guidance is essential. Therefore, it is necessary to develop a device which allow one person to acquire a form which could exert sufficient power. In this research, we developed an assist suit that assists the lower body during a forehand swing of table tennis for acquiring a forehand swing with twist of the lower body. And we conducted experiments using this suit for beginners of table tennis. As a result, the range of movement of the lower body increased by the assist suit and there are individual differences in whether to acquire a swing with twisting the lower body.

In this paper, we developed a head-rotation device for walking guidance of the visually impaired. This device utilizes the characteristic that the direction of walking changes to follow that of the head. Previous studies have confirmed that the walking direction can be controlled by giving the head a forced gyration motion. In this study, we introduced a pressure sensor to detect the user's intention of head rotation, and improved the system to allow free head rotation. By doing so, we confirmed that a more natural and reliable walking guidance can be achieved by linking the control that allows free head movement with the forced rotation movement.

By experiments using vibration stimulation to shoulder extensors (Rear deltoid，Teres major，Long head of triceps brachii，Latissimus dorsi，Infraspinatus), this paper identified the muscles involved in the illusion of shoulder joint flexion and investigated the optimal conditions for generating the illusion of shoulder joint flexion. In this study, it was found that the posterior origin of the deltoid muscle is one of the shoulder extensors most involved in the generation of the illusion of shoulder joint flexion. When vibration stimulation with an amplitude of 1.0 mm, 1.5 mm and a frequency of 60 Hz was applied to the posterior origin of the deltoid muscle, the shoulder joint flexion illusion was increased.

Lower back pain is the most serious problem for workers in various fields and therefore sophisticated and simple solution are required. As one solution, we have been developing Muscle Suit using pneumatic artificial muscle. Also, some elderly people and women want a device that is lighter and smaller than Muscle Suit. However, due to the size of artificial muscle, there is a limit to the miniaturization of Muscle Suit. Therefore, we have developed a new lightweight and compact device using gas springs, and made practical improvements based on the problem of the Muscle Suit. This paper shows the functions and mechanisms of this device and some specific design improvements.

To reduce the metabolic cost of walking and running, a walking assist system in the form of an exoskeleton attached to the ankle joint will be developed. In this study, the effects on the lower limb muscles during walking with and without the assist system to be developed will be analyzed. Joint angles measured by motion capture and muscle activity measured by electromyography (EMG) will be used for the analysis. The results of the analysis showed that the assist system had no effect at low speeds, but above a certain speed, the exertion of force was improved in some muscles. In addition, wearing the assist system decreased the angle of the ankle joint and increased the angle of the hip joint.

In this study, we aimed to develop a sound radiator driven by SMA actuators, which have fewer shape limitations. For this purpose, we developed radiators with different shapes and different ways of incorporating SMA actuators. Then, we conducted measurement tests in an acoustic measurement environment suitable for the radiators. As a result, it was confirmed that the response of a radiator as large as a headphone could be improved by increasing the number of times the SMA actuator was sewn in and by adding a taper instead of a cylindrical shape. It was also confirmed that further miniaturization will be possible.

It is desirable to use a 25G needle in puncture surgery. However, it is very difficult and robotic interventions have been attempted. The current robot has an angle limitation for puncture with a 25G needle. This is because special control is needed to suppress the deflection of the 25G needle, which is prone to bending, and the actuator needed for the control must not be in the CT imaging range. In this paper, we have developed a new transmission mechanism necessary for the robot to perform puncture without angular constraint with a 25G needle. In addition, we developed a puncture robot with this mechanism and verified its performance.

We have developed the MR fluid clutches for Fine Haptics (H-MRC) and utilized them in the Haptic MR Device for Endoscopic Surgery (H-MRD for ES). To improve its operability, we redesigned the H-MRC to reduce its size and wight. As a result, we successfully reduced by 150 g of the total weight of the operation part of the H-MRD. In addition, the self-wight compensation was conducted for its operational part. Their effects were assessed by eight healthy males and females with a pick-and-place task. The result shows significant differences in some scores of NASA-TLX and successful improvement on the skills of operators.

Personalized bone and cartilage are necessary for the reconstruction of head parts such as the ears and nose. Such fine parts can be made using 3D printing technology, and clinical applications are expected. One of the methods for use in clinical application is to form parts in an isolator in a hospital. Therefore, we are developing a small, dust-free arm-type 3D printer that can work inside an isolator. In this paper, we analyze the vibration characteristics of the arm-type 3D printer. The effects of assembling an arm-type 3D printer and a damper for vibration suppression, as well as comparisons with two general belt-driven 3D printers, are discussed.

In this paper, we applied reinforcement learning algorithm to automated counter traction control in surgery. Counter traction is a surgical technique where the forceps are manipulated so that tissue in surgical field is visible and positioned suitable for operation. Since this surgical procedure is performed repeatedly during surgery, automation is expected to reduce the burden on the surgeon. We incorporated position-based dynamics of point clouds simulating two-dimensional membrane tissue and reward function expressing optimal shape of tissue in terms of visibility and flatness with a reinforcement learning model that learns the optimal traction direction. The model was trained in a simulation environment to generate movement of forceps tips to realize counter traction. The results suggest that the desired manipulation could be acquired under several fixed conditions.

An image quality of a medical US (ultrasound) is operator dependent because a manipulation of an US probe requires skill. To support the probe manipulation, this paper proposed a novel posture detection of an US probe with a wire-grid structure in an US coupler gel. The proposed method detects the posture based on an acquired US image of the wire-grid which is fixed in the coupler gel on a patient body. Since the acquired image includes projections of wires and they are determined by the probe posture, the probe posture can be detected from the projections. Experiments evaluated the accuracy of the proposed method and showed the method achieved an easy-to-use and precise posture detection of US probe.

In this study, we use an stretch sensor to measure the stretch of the belt. We propose a method of estimating and correcting the force with which the muscle stiffness sensor is pressed by the belt from the stretch of the belt. In the evaluation experiment, we evaluate the similarity between EMG and muscle stiffness using the correlation coefficient and RMSE. The similarity between EMG and muscle stiffness was improved by correcting the external force by the belt with the proposed method.

In usual logarithmic spiral scissors, the cutting input force depends on the position of the blade when the thickness of the cutting material is uniform. We have developed a motion conversion mechanism for logarithmic spiral scissors which equalizes the cutting input force. This paper describes the experimental results of cutting dummy organs. The final objective of this research is whether the proposed scissors can detect tumors when cutting dummy organs. In the ex vivo experiments, we confirmed the maximum cutting input force was 9.0 N when cutting chicken breast as a dummy organ. The experimental results show that the cutting input force of the dummy organ was higher than that of Kent paper (2.9 N).

With the collaboration of locally operated small surgical robot in a sterile area, a surgeon can perform robotically assisted laparoscopic solo surgery. A new operation interface based on image recognition of forceps performed by the surgeon to control a surgical assistant manipulator has been developed. The forceps shaft captured using RGB and depth images is extracted by Mask R-CNN, and the orientation of the shaft is estimated for the control command. The prediction on test images of actual operation for the Mask R-CNN trained model on simulated operation images was evaluated.

By introducing locally operated surgical assistant robots in a sterilized area, a surgeon can perform safe and accurate robotically assisted endoscopic ear surgery. A new locally operated endoscope manipulator with dual linear motion that can be used by the surgeon’s left hand during endoscopic ear surgery has been developed. The linear motion of the endoscope is driven by selectable two motors. An ultrasonic motor using a rack and pinion with a linear guide leads to precise and safe motion in the ear canal. A gearless DC motor using a belt pulley with the shared linear guide leads to high speed motion outside. The dimensions of the linear motion device are 155 mm * 282 mm * 77 mm.