This study introduces a novel approach to enhance the control of cyborg insect locomotion through the utilization of a wearable UV helmet. The proposed technology leverages ultraviolet light to influence and direct the movement of cyborg insects, providing a simple yet effective means of steering. The wearable Ultraviolet (UV) helmet serves as a remote interface, allowing operators to manipulate insect behavior for various applications, such as surveillance or environmental monitoring. Preliminary experiments demonstrate promising results in steering the cyborg insects along desired paths. This research not only contributes to the field of bio-hybrid systems but also opens new avenues for practical applications in controlled insect mobility.

Molecular motors as a microrobots possess properties such as superior scalability, self-organization, and efficient energy conversion, making them potential research directions for the development of microrobots. Additionally, biomolecular artificial muscle, comprised of molecular motors and mechanical components, has garnered significant attention worldwide due to its productive fabrication process and flexible control ability. However, inherent restrictions pose challenges in achieving multi-motion capabilities, three-dimensional contractility, and higher force generation in biomolecular artificial muscle. This research focuses on improving force generation through microtubules in biomolecular artificial muscle and aims to develop a mechanical evaluation system to estimate their structural arrangement. Tetramethylrhodamine (TRITC) was labeled on the microtubules and observed in a fabricated PDMS microchannel. In this paper, we report and develop the experiment setups that observe the dynamics of microtubules in the microchannel.

Insect cyborgs are hybrid robots that combine life and machinery, and possess both biological intelligence (i.e., superior sensing, autonomous decision-making, and locomotor abilities) of the base insect and mechanical intelligence of electronic devices. In order to achieve high control performance and persistence while maintaining the ability of the base insect, it is essential to evaluate the effects of control stimuli on the base insect and to optimize the stimuli.

We propose to evaluate the electric stimulus response of insect cyborgs by measuring neural responses. The results of this study show that insect cyborgs exhibit strong neural responses at specific frequencies during electrical stimulation. In the future, the stimulation technique using these frequencies may be improved..

Previous studies implemented manual or autonomous navigation for cyborg cockroaches under an unobstructed free environment and simple obstacles. Most autonomous navigation applies an external motion caption system. This study proposes simple feedback control for cyborg insect navigation from onboard sensors such as inertial measuring unit (IMU) and distance sensor. Obstacle avoidance and autonomous navigation strategy based on onboard sensors will be applied to a cyborg insect in an unstructured environment. The experimental result showed that the cyborg insect can be navigated automatically to avoid all the obstacles in front of it.

While the automation of overhead cranes is progressing, the lift-off process is still manually operated. In this paper, we developed a pulley position estimation system for the lift-off process, aiming for stable estimation independent of the environment. The system obtains RGB images and depth information from a depth camera, which can be easily installed on actual equipment, uses semantic segmentation to determine the pully region from the RGB images, and extracts only the depth information of the pulley region. The experimental results using an overhead crane show that the proposed system can estimate a pulley position and the amount of trolley movement, even with unlearned images.

To achieve disassembly planning for large home appliances, we propose to simultaneously plan not only two-finger grasp, suction grasp, and contact by multiple end-effectors, but also the trajectory of a robot arm and rotational motion of an automatic rotation stage. This study assumes the combined use of multiple hardware devices to manipulate various shapes for the disassembly of large home appliances. To reduce the search range in the simultaneous planning, we propose the use of intermediate points for the trajectory and prioritized exploration of rotation angles of the rotation stage according to the disassembly task. Our experimental results demonstrated the effectiveness of the proposed methods by validating the generated plans in simulations and their executions in the real world.

By sharing and integrating the robot’s work experience among multiple networked locations, the project aims to build a food handling system that can flexibly respond to various environments and food products. In this project, we developed a granular food quantitative handling system based on an image recognition model using ROS2. The system was verified by actually handling granular green onions, by controlling the amount of handling depending on whether the image recognition model is used or not. The system is also configured to accommodate robot arms of different manufacturers and structures. As a preliminary step for multi-site operation, the system was actually applied to different robot arms to demonstrate and verify the effectiveness of the system. This system enables flexible operation by sharing learning data even in different environments. Further development of this system is expected to contribute to improving efficiency and reducing labor shortages in the food industry.

In production processes, there are works to collect parts from parts warehouses and parts shelves. Automation of this work is required. When the necessary parts are stored in parts boxes, the boxes must be detected and taken out from the shelves. In this paper, we use keypoint detection, a method based on deep learning, to detect grasping positions of boxes using an RGB-D camera. We set keypoints on the boxes for the learning. The grasping positions of boxes are detected by the keypoint detection model. The accuracy of this method is verified by grasping experiments.

At aircraft production sites, a wide variety of parts is handled. To prevent mistakes in parts, the parts to be assembled are checked by fitting into molds of parts kits. The aim of this paper is to automate this work using robots. The size between the molds and the parts is tight fit, so precision work is required. In this paper, we aim to fit model-less parts into parts kits using image processing. Template matching is used to obtain the position and angle of the parts and the corresponding molds from multiple molds in the kit. Deep learning is used to estimate the grasping position and reposition the part. To correct the misalignment of the fitting position, we use a sub-camera attached to five axes of the robot. An experimental system is constructed, and the effectiveness of the proposed method is evaluated.

During drilling holes to aircraft hydraulic components, burrs occur at intersections of the holes. To prevent stress concentration and flow obstruction, deburring and chamfering are performed. These tasks are currently carried out manually by skilled workers. To reduce the workload and enhance production efficiency, automation of the entire process is required. In this paper, we focused on developing a system to automate the gripping and movement of a metal part, and insertion of a borescopes into small holes. Position detection of the part was performed by using rectangle detection. The gripping and movement of the part was carried out using a robotic hand based on the acquired position information. In the insertion experiments conducted on three blind holes and one through hole, the insertion success rate was 96% for all holes. Due to some manual processes involved, complete automation was not achieved.

This paper describes robotic assembly of gearmotors. A gear motor consists of many parts such as a rotor, a stator, an anti-load bracket, a gear box, etc. Assembly task of the rotor, stator and anti-load bracket into the gear box is discussed. We use a robotic arm equipped with a force sensor, an RGB-D camera, and a robotic hand. Another RGB-D camera is placed in the side of the gear box. Depth information is used to separate background information from a captured image. In order to recognize the assemblies, keypoint matching and template matching are used. We make contact judgment and position correction by referring changes of the force date.

In this paper, we discuss automatic assembly of gearmotors. The gearmotors consist of multiple parts. To recognize the shape of the parts and screw holes, we utilize semantic segmentation. To detect angle of the parts, we use position of the screw holes. Automatic assembly is performed based on the recognition and detection. Through experiments, it is validated that our proposed method is robust to variations in brightness.

Deep imitation learning is a technology wherein robots autonomously learn actions by directly instructing them from humans. By incorporating image information into this learning process, automatic generation of grasping actions for robot picking solely from images is possible. Because it is expected to be influenced significantly by the visual appearance of the images, we investigated and identified the camera position that yielded the highest picking performance by conducting deep imitation learning at various camera positions. However, the analysis of why differences in results occurred was insufficient. Therefore, in this study, we analyze by visualizing the regions of interest of a CNN using Grad-CAM.

Using a camera attached to the tip of the robot to measure the position and orientation of markers attached to the environment enables a manipulation that is independent of errors in the installation position of the robot arm. However, the measured marker position includes measurement errors, and they propagate to cause task position estimation errors. This paper shows how to calculate approximate task position estimation errors when measurement errors were given by a known normal distribution. The effect of the approximation error is theoretically discussed. In addition, a numerical experiment confirmed that practical results with a small amount of computation and small approximation errors can be obtained by the approximated calculation.

TIIT conducted research and obtained a patent for a random picking robot system characterized by the use of visual and tactile sensors, which has advantages over existing systems in terms of price and time required to change parts to be grasped. In this paper, we report on a robot hand equipped with a tactile sensor that has a special mechanism to adjust the gripping surface, a method to correct object recognition errors based on visual information during bulk picking, and a high-precision picking method that incorporates a search motion equivalent to groping while monitoring the gripping state, which was studied jointly with a company to commercialize the research results.

This paper discusses assembly operations using a single manipulator and a parallel gripper to simultaneously grasp multiple objects and hold the group of temporarily assembled objects. Multiple robots and jigs generally operate assembly tasks by constraining the target objects mechanically or geometrically to prevent them from moving. It is necessary to analyze the physical interaction between the objects for such constraints to achieve the tasks with a single gripper. In this paper, we focus on assembling pipe joints as an example and discuss constraining the motion of the objects. Our demonstration shows that a simple tool can facilitate holding multiple objects with a single gripper.

We are working on automating the attachment of candle decorative stickers using a robot arm. We investigated in detail the situation of misalignment of the affixing position caused by the prototype system by image processing furthermore, we examined whether it can be applied to the automation of the determination of good/defective products.

Welding robots can achieve high-quality and highly efficient welding without the need for a skilled welder by selecting appropriate welding conditions and controls. However, automated welding has not yet been able to achieve bead formation with the same level of aesthetics as skilled welders who use weaving motion. This study conducted welding experiments with various weaving motions by changing four speed parameters: feed rate, amplitude size, frequency, and rest time. An analysis was conducted to examine the relationship between the dimensions of the beads obtained and the weaving motion. These data were used to develop an automatic compensation system for feedback welding.

We proposed a method for measuring the position and orientation of cable using an RGB-D camera aiming to enable the automatic control of a multi-rotor flying robot equipped with a power line docking capability and active translatory drive system. In cable detection, a trigger signal for starting to perform UAV ascent and yaw rotation movementsis generated when the detected cable, characterized by the maximum area of black color, meets specific conditions regarding their position and orientation. Through image processing experiments, the feasibility of estimating the position and orientation of the power lines was confirmed.

In this paper, we focus not only on the approach of a small drone when flying close to a wall, which is the surface to be inspected, but also on the influence of the drone's own wind on its stability when approaching the ceiling, left and right sides, respectively. We then examined the evaluation method.

In this paper, we examine the rapid ascent capability of a dual propulsion VTOL (Vertical Take-Off and Landing) drone based on its dynamic model. Through numerical optimization, we find the rapid ascent motion that maximizes altitude change within a specified flight duration. The optimization results show that employing both the rotors for rotary-wing mode and the rotor for fixed-wing mode allows for a greater ascent compared to conventional fixed-wing mode flight.

Recent developments in drone technology have expanded their use in aerial filming. However, capturing footage over large areas remains a challenge, requiring advanced skills and significant labor. This study proposes a straightforward method to automate the process of creating datasets from drone-captured videos. Our approach enables drones to operate based on natural language instructions, focusing on overcoming the previously identified challenge of generating such instructions. By integrating image recognition with Large Language Models (LLM), we provide a practical solution to enhance the automation of drone filming. This method simplifies the dataset creation process, making it more accessible for tasks requiring aerial footage, and offers a step towards more intuitive drone control in various applications.

This paper proposes a thrust estimation method based on audio information utilizing a regression Convolutional Neural Network (CNN). This paper conducts a fundamental experiment involving the force estimation of a single rotor fixed to the ground. The estimation performance in the proposed method was almost the same as in the previous study, in which a force-sensing resistor is integrated for the thrust measurement mechanism.

In recent years, Open Source Software (OSS) has played an essential role in the development of Unmanned Aerial Vehicles (UAVs) and other robots. However, there is a risk that robots developed using OSS may be used for inhumane purposes, such as terrorist acts unintended by OSS developers. This paper describes verifying the functionality at the processor level to prevent the abuse of UAVs by seizing control. In this paper, a prototype of a secure processor is developed by using FPGAs, that cannot be modified by the user. Besides, this paper demonstrates fundamental verification to assess the processor’s ability to detect intrusions into restricted areas and to execute forced soft-landing procedures.

Airflow analysis is necessary for safe veranda delivery by UAV. This paper describes a two-dimensional airflow analysis of an enclosed space with an opening using Physics-Informed Neural Networks (PINNs). The performance of real-time airflow analysis and accuracy evaluation are carried out by comparing those with the finite difference method.

Aging infrastructure facilities are increasing in Japan. Therefore, there is a growing demand for infrastructure inspections conducted by drones, which are manpower-saving and highly maneuverable. However, there is no standardized method for quantitatively evaluating drone inspection capabilities. In the previous paper, a method was proposed to evaluate drones’ inspection capability. Furthermore, drones’ performance has been evaluated quantitatively by making and inspecting a test specimen of cracks. In this paper, a crack detection experiment was conducted using a test specimen of cracks. The obtained results evaluated the relationship between the drone’s flight parameters (acceleration in the x-direction, roll angle and pitch angle), inspection time, and the inspection success rate evaluated and discussed concerning ”drone flight and control performance.”

Recently, robot teleoperation has been researched. This is suitable for performing tasks in environments that are inaccessible to humans. In particular, remote control of aerial robots is used for high-place work. Currently, mainly used aerial robots are multirotors with manipulators attached to them. Such robots can replace humans in high-place work. However, they have not been able to approach locations in a cluttered environment. One solution to this problem is to use an articulated aerial robot. Articulated aerial robots have redundant Degrees-of-Freedom(DOFs), squeeze in the cluttered environment by changing the shape and approach the end effector. To remotely control an articulated aerial robot, it is necessary to control the robot’s multiple DOFs, such as joints and positions. However, it is difficult for conventional devices to simultaneously control these DOFs. A device suitable for remote control of articulated aerial robots is needed. To solve this problem, we propose a floating-based device with a skeleton similar to that of an articulated aerial robot. Beside, some tasks require the robot to exert force with its end effector, and external forces such as the reaction force and friction from the environment can make the robot’s flight unstable. It is desirable for the robot to control forces autonomously and stabilize its flight. Only using the thrust scalar to compensate for the external forces has the limitation, thereby we also focus on the thrust direction, and introduce a vectoring apparatus that can change the direction. In this study, we propose an autonomous control method that achieves both end effector force exertion and flight stabilization by controlling the vectorable thrust force. Finally, we verify the proposed methods by conducting wall-cleaning experiments using an articulated aerial robot with vectoring.

In recent years, multimodal locomotion in robotics field has been investigated intensively. Although aerial robots have three dimensional locomotion ability, its energy consumption is large and flight time is not long. To address this problem, multirotor robot platform with rolling locomotion ability in terrestrial domain was proposed. In this work, we show a thrust control method for the ground motions of such robots. Then, we introduce the implemented prototype and verify the proposed control method through experiments.

The purpose of this study is to develop a wearable walking assist robot that can be used in daily life. Gait stability decreases when encountering various obstacles during walking. Therefore, it is necessary to quantify the influence of assistance on users’ gait stability when encountering disturbances while walking, and to develop stability evaluation indices. This study focuses on turning gait assistance and investigates the effect of torque assist of wearable robot on the stability of users’ turning gait. The assist patterns used in the experiment are one based on the angular velocity of the lower limb joints during walking and the other based on internal torque. The results of experiments suggested that assist based on angular velocity has the highest stability and is the optimal assist for turning gait assistance.

This paper proposes a wearable motion classification system using commercially-available conductive threads and everyday garments．The unique feature of the proposed system is a sewing-type bending sensor that does not necessitate tight-fitting with the body，which is a sharp contrast with the existing motion sensing system．This simplifies integration by allowing motion classification even onto loose-fitting everyday garments．The sensor that shows a change in resistance to bending deformation is realized by applying multiple zigzag stitch to the fabric．We designed this stitch to categorize motion without making joint angle measurements. This sensor exhibits a large resistance change during the phase transition of motions (e.g.,stance and swing phase)．Applying fast Fourier transform to the data of these sensors reveals motion magnitude and period，enabling motion classification．The prototype made of regular pants successfully demonstrated distinguishing four motions: walking，limping，and stair climbing．

In this study, we developed a meal support system using a wearable mechatronic finger. The finger device was proposed to support a daily grasping task in our previous study and comprised two fingers opposite each other. The present study considers the use of chopsticks with this finger device. We designed an attachment that installs the chopsticks to the device and switches the device function mechanically from grasping to meal support. The chopsticks then contact with the ball of fingers, where a pressure sensor sheet is installed to feedback on the reaction force from the target food. Additionally, an object recognition-based control system is employed to adjust the chopstick motion to the target food. We conducted evaluation experiments of the proposed device with a commercial bento and confirmed that the contents could be picked up with a few kinds of the servo command on the closure angle of the chopsticks.

In our previous study, an elbow support power assist suit for underwater work has been developed. In this study, a shoulder support power assist suit for underwater work was developed so as to extend an assist function of the underwater assist suit. In order to evaluate an assist effect of the developed power assist suit, experiments of the movements such as rolling, carrying, and holding a heavy object, were conducted to measure surface electromyograms of the biceps brachii muscles. As a result, a reduction of the burden on the biceps brachii muscles under wearing the shoulder support power assist suit for underwater work was confirmed. Thus, the effectiveness of the developed power assist suit was verified.

A wearable sensor for measuring core body temperature is expected to be developed as a helpful tool in preventing heat-related illnesses. One potential method is the heat flux method. However, this method has technical challenges, including measurement errors caused by changes in ambient temperature and air flow. We have recently developed a new method to address this problem. However, achieving both high accuracy and low weight of the measurement probe remained a practical issue for wearable sensing. Here, we focused on the probe cover, accounting for most of the probe's weight, and investigated the relationship between its shape and accuracy. Our numerical simulation analysis revealed two types of probe cover structures that can achieve both high accuracy and low weight. Simulation experiments with the prototypes demonstrated these are more accurate than the conventional probe. These findings will contribute to the advancement of measurement technology for wearable core body temperature sensing.

Mastering piano performance requires learning various techniques such as precise keystrokes, appropriate fingering, rhythm, and dynamics, which can take considerable time and effort. This research proposes a system to assist in improving piano performance by physically aiding the positioning and timing of keystrokes. The method involves applying functional electrical stimulation to the back of the performer’s hand to support fingering and keystroke timing, and moving the performer’s arm to the correct position using a sliding mechanism and motor. Experiments showed the system effectively reduced keystroke timing delays by approximately 50% and enhanced musical performance.

One method of assisting workers is the use of lumbar assist devices. Various types of lumbar assist devices have been researched and developed by various research institutes, and their effectiveness has been demonstrated. However, there are few studies on the effects of lumbar assist devices on the body. In particular, changes in compression force and shear force applied to the intervertebral disc may increase vertebral displacement and cause low back pain. In this study, the erector spine muscular strength, the compression force of the intervertebral disc and the shear force of the intervertebral disc were calculated when a lumbar assist device was used, based on a lumbar load evaluation model. In addition, a lumbar assist device that took into account the load on the intervertebral disc was discussed.

This research aims to develop a small, cane-type robot that is easy to use in narrow indoor spaces and does not require restraints to assist the elderly in standing-up, walking, and sitting down. In this study, we focused on assisting in standing-up and analyzed standing-up motion using a cane to determine effective robot movements. We measured knee joint moment values using a cane-type measuring device we developed and compared different conditions of cane tip position and cane length. The results showed that positioning the cane tip most forward and on the right side, and using the highest cane length, reduced the burden on the lower limb the most.

It is important to detect anomalies of the user for a robot to assist an elderly person to stand up. A calculation method to calculate the user's candidate center of gravity with a small number of sensors have proposed for anomaly detection. However, the anomaly detection was based on two-class discrimination using machine learning, so it was required to measure anomaly data in addition to normal data. Therefore, this study analyzes the trend of the candidate centers of gravity during anomaly standing for anomaly detection without requiring anomaly data. The analysis showed that the maximum value in the y-direction and the minimum value in the z-direction of the candidate centers of gravity changed significantly with the abnormal standing. By setting a threshold for the first-order derivative of each value, it was possible to detect abnormal standing correctly.

The demand for personal vehicles (PVs) is increasing with the aging of society, and PVs are expected to become a means of transportation for the elderly. However, the operation of PVs by the elderly is fraught with risks such as collisions. In response to this problem, MR-based maneuvering assistance and training systems have been developed. However, since these systems are based on manual operation, they cannot guarantee driving safety. In this paper, a brain-computer interface (BCI) using steady state visual evoked potentials (SSVEP) to develop an automated driving system is proposed that detects the intention of the PV user and uses it into driving for safe driving.

In recent years, the spread of coronavirus has led to an increase in chair-sitting opportunities, such as telecommuting and online classes. As a result, sitting posture has worsened, causing many health problems such as gait failure and diabetes. Therefore, the purpose of this study is to develop a wire array sensor that can analyze and correct sitting posture in order to prevent health problems. And in this paper, a system created with four motors and wires was used as a preliminary step for heat mapping and COP position estimation with the wire array sensor. As a result, it was confirmed that accurate force estimation was possible

Insomnia is a growing problem, especially in developed countries.There are solutions using drugs, but they have side effects. Therefore, we are considering the use of vibration stimulation as a kind of relaxation method, but there is no research on these methods, so we investigated the effects of vibration amplitude and stimulation on sleep onset.We investigated the effects of vibration amplitude and stimulation on falling asleep.In Experiment 1, we investigated the subjective evaluation of vibration stimuli when falling asleep.In Experiment 2, nine vibration conditions were presented for 10 minutes, combining the resting state, amplitude, and frequency.The results of Experiment 1 showed that the range of perception was narrower in the 40-100 Hz interval, and that the error range was narrower.The results of Experiment 2 showed that vibration had a relaxing effect, but there were individual differences among subjects.

In general, when machines assist human movements, the importance of elements where humans move themselves is emphasized. This helps actively engage residual muscles, thus preventing their deterioration, leading to improvements in rehabilitation and quality of life (QoL). This study proposes a control method for a planar 2-degree-of-freedom robot, based on the Assist-as-Needed approach, to provide minimal assistance to a diverse range of individuals during the crucial standing-up motion in daily life. Focusing on the pivotal joints of the knee and hip during this motion, the proposed method observes the posture and hand force of individuals to determine which joints require assistance. Subsequently, simulations are conducted to provide support during the standing-up motion by adjusting the assistance ratio between the knee and hip joints dynamically.

It is hard works to move heavy loads for health person and even more for elderly people. Carriages and walkers are generally utilized in order to reduce the burden, however it is difficult to move these devices at uneven area and they take much space. Therefore, wearable devices which can perform bipedal locomotion are desired. Because the existing devices adopt electric motor and pneumatic actuator, they have problems about short acting time of the battery and air cable connection. So we have developed a simple wearable device with low power mechanism such as disc brake or ratchet. This device has a unique point which prismatic joint for the knee is adopted.

We have been studying a sit-to-stand assistance device that does not use external power and is safely usable at home. The device is like a cane with a rod to push one's waist. The bottom of the cane forms a non-linear rolling shape designed to push the person's waist to rise in a desirable form. Also, the cane is designed to roll forward by pulling a lever attached to the cane via linkage. By this mechanism, a user can naturally move his/her body forward by the pulling action while his/her waist is assisted by a rod attached to the cane. In our previous reports, the waist assist route was taken from healthy persons' stand-up motions without using handrails. However, the route may differ when the persons use handrails while standing up, which may cause the rolling cane's bottom to be re-designed. This report shows the waist route measurement while standing up with various heights of handrails. We found that the handrail 80cm high at the beginning and 70cm high at the end showed better stability and waist route, which contributed to reducing the size of the device.

We have been investigating a method to operate an electric prosthetic hand by measuring swelling or hardness changes of muscle with a soft-body pressure sensor instead of myoelectric potential. We found that this method is much more stable and easier to operate than myoelectric potential. In addition, it has been shown that it is possible to grasp silken tofu (very soft and fragile food) with an electric prosthetic hand by attaching a soft pressure sensor to the fingertip of the prosthetic hand. We also developed a sensor with three photo-reflectors embedded in a hemispherical balloon-shaped silicone rubber to measure not only the pressure but also the direction of the pressure. In this report, we newly developed fingertip sensors for the fingers of the prosthetic hand. It includes 7 photo reflector elements (4 for 4 soft pressure sensors and 3 for one direction indicator) on the nail and it can measure the attaching position of the finger pad, shear displacement of the finger pad and the compression force of the nail. By using this fingertip sensors, we can pick soft material and we can also break hard objects if necessary. We succeeded in peeling off the shell of a boiled egg by these fingertip sensors installed prosthetic hand.

We have developed an intraoral input device using tongue. This device contains three photo-reflectors in a balloon of silicone rubber and it can measure the quantity and direction of the pushing force of silicone rubber film. By using this device, we can make a controller that generates information by the tongue. We have applied this device to control a robotic arm for a third arm for soldering workers, to control an electric wheelchair for the physically challenged, and to control a VR device. In all these applications, we have achieved very good operability. However, there is a drawback that we cannot speak while using this device. In this paper, we developed a speech enabled intraoral input device. The sensor was miniaturized to reduce the inhibition of tongue movement. In addition, the mouthpiece was improved to allow the voice to pass through, and at the same time, we worked on maintaining operability. As a result, we were able to develop an intraoral input device with good utterance and operability.

In a previous study, we developed a treadmill-type walking device with applying a wire traction force and confirmed to improve the walking experience by offsetting the inertial force when the device was driven. However, the method of applying traction force has not been fully investigated. This study develops trunk equipment for effective and comfortable traction force application. It consists of an arm-ring structure where the traction wire is attached to the ring, connected to the trunk equipment with a V-shaped arm, a rigid plate not deforming with the force transmitted from the arm, and a sponge absorbing the gap from the plate to the body surface and individual differences. The results of walking experiments indicated that the proposed equipment improved the walking posture, and the questionnaire confirmed that the natural walking experience was improved.

Exercise-related diseases are increasing in Japan due to the aging of the population. In this study, we focused on osteoarthritis of the knee, one of the most common movement diseases, and examined the burden placed on the knee by the foot arch. Since this disease causes pain when ascending and descending stairs and walking, we fabricated a foot with a function to reduce the impact on the ground like a human foot, and verified the effect of the height of the foot arch on the force applied to the knee when the sole of the foot contacts the ground. We also evaluated the effect of the foot on the robot's movement by attaching the foot to the robot.

Adaptation, learning and plasticity of visuomotor transformation are a neural strategy to update and enhance an internal process for the new environment. This study presents the formation, collapse, and reformation of patterns in the error-action relation in the error phase space during nine days of visuomotor learning that involves a participant who performs manual tracking in a heuristic trial-and-error way under reversed/unreversed vision. The analysis results show the geometrical features of dynamic structure in the error phase space reflect the level of individual ability and performance, suggesting the potential for quantitative evaluation of motor planning.

Humans are able to walk by controlling multiple joints simultaneously. In this study, we estimated the impedance of ankle joint with sudden stops of the treadmill belt.We tried to reveal the viscoelastic properties of the lower limb joints considering the interlocking of multiple joints during walking. It was shown that the stiffness of ankle joint has an increasing tendency from 15% to 35% of the walking cycle, and that it can be expected to reach its maximum between 35% to 40% of the walking cycle. We confirmed that the muscle activity level of the gastrocnemius muscle, which is mainly responsible for plantar flexion during walking, roughly corresponds to the phase when the stiffness is expected to be maximum. It was shown that lower limb impedance estimation using sudden stopping of a treadmill is considered to be generally effective in estimating stiffness.

The present study focuses on developing a technology to support individuals with prolonged sedentary behavior and low motivation for physical activity. The paper introduces a device designed to induce voluntary seated pedaling movements and assist the user in moving their legs with less conscious focus.