In Chemical Mechanical Planarization (CMP) technology, the CMP pad conditioner plays an important role in stabilizing the removal rate by scraping and roughening the pad surface finely. In this study, we developed a fiber conditioner that can perform conditioning based on the pad surface shape in order to stabilize the removal rate. Fiber conditioners have a relatively smaller rake angle than conventional diamond conditioners, so they are better at fine conditioning than conventional diamond conditioner. In this report, we developed a method to visualize contact area in the pad surface based on light reflection and scattering using a coaxial and ring light for the purpose of evaluating the pad surface condition in fiber conditioners. New methods of visualization of the 3D finely scraped pad surfaces provide the prospect of clearly identifying the differences in pad surface condition dependent on different conditioners.

Various methods of 3D gel printers have been reported to produce on-demand flexible structures such as 3D cell scaffold materials and soft robots.3D gel printers can produce hollow structures of arbitrary shapes, which is difficult to achieve by mold forming.On the other hand, most 3D gel printers to date have been based on a single method, and it has been difficult to incorporate multiple gel materials with different mechanical properties into a single model.Therefore, we have developed a 3D gel printer (GelPiPer-multi) that employs two methods of modeling: a free surface method in which a laser beam is irradiated from the top surface of a tank filled with a solution, and a material extrusion method in which material is ejected from a nozzle.In this study, we report on the GelPiPer-multi and its modeling product, a composite gel structure.

In the grinding process, dressing significantly changes the grinding wheel surface condition and has a significant effect on the grinding resistance and surface roughness. In other words, proper dressing of the grinding wheel surface is important for good grinding. In this study, dressing conditions were determined by changing the dressing overlap ratio, and dressing was applied by one-way and reciprocating dressing, and by single-point and prismatic dressers. The effects of dressing overlap ratio, dressing method, and dresser shape on grinding characteristics and grinding wheel wear were evaluated. The experimental results showed that when the dressing overlap ratio was greater than 2, grinding characteristics and grinding wheel wear did not change because the grinding wheel surface condition did not change.

Recently，soft robotics research has been conducted on soft-bodied organisms（squid，octopus，jellyfish），and Inter-Crosslinking Network（ICN）gels with high water content have been studied to provide flexibility and high fracture strain．The 3D gel printer technology has evolved，and technologies to add deformation response to 3D printed gel materials have been studied．One of the most promising methods is to add magnetic field responsiveness by dispersing magnetic particles．In this study，we analyze the molding characteristics of magnetic ICN gel using a 3D printer with a regulated liquid surface system．In particular，we will clarify the parameters that contribute to 3D printing．

Metal bone fixation devices used for fracture treatment have a higher elastic modulus than that of autogenous bone. However , this high modulus weakes the surrounding bone. Therefore, we focused on a composite material of poly(lactic acid) (PLA) and tricalcium phosphate (TCP), because a bioabsorbable material with an elastic modulus similar to that of natural bone. However, apprication of TCP/PLA composite is difficult because of lower strength. Therefore, we focused on drawing to improve the strength. However, it is difficult to form complex shape such as screw. In this study, TCP/PLA screws were molded by hot forging, and the strength of this screws are eveluated for shear and torsion tests. The results of shear test showed that shear strength increased extrusion ratio (ER) in hot forging. The shear strength decreased with increasing TCP content. On the other hand, the shear strength of the TCP/PLA screw increased with hot forging, and the strength at ER8 and TCP content 15 mass% achieved higher strength than that of without forging and 0 mass%. Torsion test results indicated that the addition of TCP inhibited axial orientation. The maximum torsional moment increased with TCP content.

Forged steel often shows a fibrous microstructure called “metal-flow”. It has been reported that the high toughness and strength of the steels are observed along the metal-flow^（1）. Inhomogeneous chemical composition is considered to be one of the reasons for the formation of the metal-flow⁽²⁾. However, detail relationship between mechanical property and metal-flow has not been clarified. Formation mechanism of metal-flow is also unclear. In this study, microstructural observations and tensile tests were conducted on SNC631 before forging, after forging, and after annealing. The results showed that annealing reduces the density of metal-flow. Low yield and fracture stress is observed after annealing.

The composite of tricalcium phosphate (TCP) and poly(lactic acid) (PLA) was attracted attention as a material of bone fixation device that is bioabsorbable and can decrease burden for patients. However, the mechanical properties of TCP/PLA composite are not sufficient for application as a bone fixation device. Therefore, the interface treatment between TCP and PLA has been investigated to improve the mechanical properties. In order to reduce the molding cost in the interface treatment, this study aims to reduce the amount of solvent used in the interface treatment. Changes in the weight of ethanol used as solvent during the interfacial treatment were measured. As a result, the weight of ethanol in the mixture decreased with interfacial treatment time and ethanol was almost volatilized after about 75 minutes. And, effect of the amount of stearic acid on molding and mechanical properties of TCP/PLA composite was investigated as a basic study. The viscosity of melted TCP/PLA composites interface treated with stearic acid was investigated by melt flow rate (MFR). The results showed that compositing with TCP and interface treatment reduced the melt viscosity during molding. The tensile strength of the TCP/PLA composite with 24 phc was slightly higther than that of 0 phc and 6 phc. On the other hand, the elastic modulus increased at 6 phc, and decreased at 24 phc. This increase in tensile strength and modulus might be due to the increased interfacial strength between TCP and PLA resulting from the interface treatment.

Since the waviness on the surface of optical elements has a long period behavior, it is difficult to remove it accurately and rapidly by polishing. Therefore, in-process measurement technology that provides real-time feedback on waviness measurement results to the processing machine is required. Laser inverse scattering is a measurement technique that is suitable for in-process measurement and can perform three-dimensional measurements. The intensity distribution of the diffraction pattern generated by scattered light on the sample surface in the laser inverse scattering depends on the spatial frequency of the surface topography. Therefore, the phase distribution reflecting only waviness can be obtained by filtering out the intensity distribution other than the spatial frequency component of waviness. After the phase restoration, the waviness of the object can be obtained. To develop an in-process waviness measurement method, we propose a method to measure only waviness by filtering diffraction patterns using an optical spatial-frequency filter in the laser inverse scattering. In this report, simulations were performed to measure only waviness from shapes with waviness and roughness. Since the proposed method is different from conventional filtering methods, the amplitude transfer characteristics of the proposed method were investigated and compared with those of conventional filtering methods.

Surface-activated bonding is a wafer bonding technology where the wafer surface is activated by irradiation with a fast atom beam in a vacuum. We have demonstrated that it can improve the irradiation efficiency and reduce bonding defects caused by carbon emissions by controlling the motion of charged particles with a bidirectional magnetic field applied to the beam source. In recent years, there has been a growing need for a beam source that can irradiate a larger area. In this study, we propose a prototype of a unidirectional magnetic field-applied fast atom beam source that can be lengthened or enlarged. The proposed fast atom beam source achieved more than 1.5 times higher oxide film removal performance compared to the conventionally used fast atom beam source. In a long-term irradiation experiment, the proposed fast atom beam source emitted no carbon particles even after more than twice as many irradiations as the conventional fast atom beam source.

This study aims to determine how friction-induced transferred film affects the friction properties of nanostripe structures. In this report, two types of nanostripe structures were used. One was composed of Au and Cr and the other was Ag and Cr. First, each nanostripe substrate was rubbed using a ball-on-plate friction machine. The friction coefficient of Ag-3Cr was lower than that of Au-Cr when friction tests with Al2O3 ball took place in a vacuum environment. The friction coefficient of Au-Cr decreased at one time, but then increased and did not stabilize. In contrast, the friction coefficient of Ag-3Cr was stable at 0.2 from around 200 times. Then, the wear track on the nanostripe structure was measured using conductive atomic force microscopy(c-AFM). c-AFM results showed that the current flowable region was wider for Au-Cr than for Ag-3Cr. In the future, we will investigate the influence of the transfer film on friction reduction and the distribution of film thickness by measuring friction in the wear traces using AFM.

Surface activated bonding is a three-dimensional wafer stacking technology utilized in the semiconductor field. In this process wafer surfaces are activated by argon atoms irradiated from a fast atom beam source, allowing bonding to take place is at room temperature and low pressure. However, fast atom beam sources present an issue wherein carbon agglomerates, responsible for bonding defects, are discharged alongside the argon atoms. Therefore, it becomes necessary to replace the carbon electrode before these aggregates released. However, quantitatively measuring the ideal time for replacement has been challenging. To address this concern we developed and validated a quantitative in-process measurement method for monitoring discharged carbon agglomerates. This method involves illuminatinge the wafer surface, capturing dark-field images of carbon agglomerates, and performing image processing to quantify measure the amount of discharged carbon agglomerates discharged. In order to establish a reliable relationship between the size of carbon aggregates (maximum Feret diameter) and the corresponding maximum Feret diameter in the dark-field image, we conduct an investigation. The results of the measurement indicated a significant increase in the number of discharged carbon agglomerates after 150 irradiations. Moreover, the largest number of carbon agglomerates was detected in the area between 5 and 7 pixels. These results suggest that the formation of carbon agglomerates occurs through exfolitation, particularly when the irradiation is temporarily interrrupted before the 150th irradiation.

Fast atomic beam sources used for surface-activated bonding need to uniformly irradiate a large area. In conventional methods, the irradiation characteristics of such a fast atom beam source must be improved through trial-and-error modifications of the beam aperture arrangement, among other techniques, and by conducting silicon oxide film removal experiments and verification, all of which are inefficient. In this study, we employed the PIC-MCC and DSMC plasma simulation methods to acquire the Ar flux and incident energy distributions on the wafer surface. These distributions predict the amount of oxide film removal caused by the fast atomic beam, and based on them, we constructed a model that predicts the distribution of the amount of film removed. Subsequently, we validated of the prediction model by conducting silicon oxide film removal experiments and obtaining the distributions of the amount of film removed.

Since high torsional performance is required to evaluate the durability performance of clutch dampers in automobiles, hydraulic servos with high power consumption are used in torsion testers. In this study, to reduce power consumption, we constructed a test system to evaluate the durability performance of the clutch damper using an electric servo torsion tester. In the evaluation test, it was confirmed that the torsional torque in the specific frequency range was amplified during the adjustment of the inertia on the rotating shaft. Furthermore, we constructed a control system that considers the non-linear torsional stiffness of the clutch damper and confirmed a 90% reduction in power consumption and a stable response to non-linear torsional torque through actual equipment verification.

Infrastructure facilities have been noticeably deteriorating and need to be inspected. The use of pipe-travelling robots has been investigated, but the robots developed so far are too large for smaller pipes. As the robot is downsized, however, the cables used for communication and power transmission disturb the movement of the robot. In this report, we used microwave guided modes propagating along a pipe wall as a communication method. An experimental robot platform for small-diameter pipe inspection was constructed. We confirmed the feasibility of the communication method by conducting a 3D electromagnetic field simulation and experiments on the transmission properties. In particular, the experiments on data communication and picture transmission demonstrated the possibility of communication with the robot. A small pipe-travelling robot platform for experiments on wireless communication was designed and fabricated. We confirmed that the prototype robot was able to communicate wirelessly by using microwave guided modes propagating along the pipe wall.

Optimal control of robot is desired to automate complex tasks performed by humans. For example, a picking robot system is required to perform high-speed picking motion and grasp a variety of products. To achieve this, the arm and hand motions must be coordinated. However, since there are multiple controllers in the robotic system, it is difficult to deal with this problem only by coordinating them. In this study, we propose to apply the method of controlling multiple tasks in parallel according to the robot's states to the picking robot. The proposed method checks the execution status of tasks with multiple control modules and divides tasks according to their operation. Then it makes control requests at the appropriate time. We incorporated the proposed method into the picking robot system and measured the motion time. As a result, we confirmed that the grasping motion time was reduced by 0.53 seconds (24%) compared to the method without parallel control in the coordinated motion of the arm and hand.

In this study, plastic deformation control without force information is designed and implemented. This control strategy is derived based on the concept that the shift in position and posture attributable to an external force as the deformation of the robot. The Maxwell model that describes plastic deformation is derived as the simple compressed expression using the convolution. Based on the expression, a control law that does not require force information is proposed by not performing the inertia shaping. Physical simulation with a robotic arm are executed to validate the proposed control law.

In order to achieve high-precision position tracking performances in positioning systems, the friction phenomena must be taken into consideration during the development of the control algorithm. For effective compensation, the value of the friction force must be known in every domain of operation of the mechanical systems. This study presents an approach for estimating the parameter space representation of feed drive systems with nonlinear friction. Its effectiveness is demonstrated through a numerical example of a feed drive system with nonlinear friction.

Exterior wall tiles risk delamination with time-related deterioration. To prevent exterior wall tiles from falling, a full inspection of the building must be carried out every 10 years. A visual inspection and an impact acoustic inspection by workers are used to evaluate the condition of buildings. These inspections need scaffolding, aerial work vehicles, and workers' labor costs. For residents, the cost of inspections is an excessive burden. Therefore, there is a need for a reliable and low-cost inspection method. In this study, we developed a rotating hammering device that hits the wall surface at high speed. To prevent the chipping of tiles caused by hammering, the resin was used as the hammer. The performance of the hammering device was evaluated by verifying the stability of the impacts and the performance in defect detection. As a result, we verified the effectiveness of the hammering device using a resin hammer.

This paper presents a control method for transitioning the formation shape in the leader-follower method, shape transition, as a new feature to bring flexibility to mobile robot formation control. First, we propose a basic shape transition method under formation control by the leader-follower method, and then demonstrate the effectiveness of the shape transition method proposed in this paper through control experiments using the Pioneer, a mobile robot with multiple sonar sensors.

An inertia matrix in the equations of motion for a manipulator is positive definite. The positive definiteness of the inertia matrix depends on set of base-parameter values. Once all the base-parameter values are given, each element of the inertia matrix becomes a function of the joint variables. The base parameters, which consist of the inertia parameters, in the manipulator's motion equations, need to be estimated using parameter identification. Therefore, the inertia matrix can reveal non-positive-definite postures depending on the combination of base parameter values, and these combinations are referred to as physically impossible. The positive definiteness of the inertia matrix based on base-parameter values pertains to dynamics. On the other hand, there is the concept of manipulability in kinematics. While positive definiteness and manipulability may seem like separate concepts, it was found that for a three-degree-of-freedom manipulator, there is a correlation between the postures where the inertia matrix becomes non-positive-definite due to physically impossible base-parameter values and the postures with low manipulability.

In this paper, we focus on image sensor-based line detection systems, which are necessary to handle real-world events in the digital world, and propose a monitoring system using an Event-based camera, which is a sensor with features not found in conventional cameras. To address the main challenges of noise, feature extraction, and object recognition in Event-based cameras, we introduce three novel image processing algorithms: noise removal using Spiking Neural Network (SNN), feature extraction based on topological mapping using Growing Neural Gas (GNG), and object recognition using Graph Convolution Network (GCN). Furthermore, we develop a pan-tilt mechanism using two Eventbased cameras, enabling us to switch between recognition and tracking roles and facilitating effective target tracking. Three main evaluation criteria were used for each experiment: control of the pan-tilt mechanism with two event cameras, object recognition using GCN, and whether the tracking target could be captured at the center of the field of view. Through experiments, we confirmed that the proposed method can perform real-time target tracking using event cameras and that the proposed method is effective for fast-moving objects that cannot be measured with conventional cameras.

Vibrational stimuli provided to the tendons induce the kinesthetic illusion (KI). Our goal is to control KI and apply it to rehabilitation and VR. Since the kinesthetic sensation obtained by KI is not accompanied by actual movement. That is why, this sensation is inhibited with the eyes open, so research is usually conducted with the eyes closed. In this research, we aim to develop the kinesthetic control system that feeds back the kinesthetic sensation felt by the subject with KI through the Head Mount Display (HMD) visually. The target of stimulation is the flexor carpi radialis muscle tendon of the right wrist, and the subject perceives the illusion of stretching the right wrist when this tendon is stimulated. The subject is made to reproduce this right wrist sensation on the left wrist, this left wrist movement is measured with the motion capture and the sensation reproduces it as the movement of the right hand through the HMD. Through this system, we try to clarify the combination effect of KI and visual feedback.

We are aiming to develop a nail color tactile sensor that can be easily attached to human fingertips to accurately record tactile information. The sensor will estimate both vertical and shear forces applied to the fingertip using machine learning algorithms based on changes in nail color. Previous prototypes have encountered challenges due to relative displacement between the sensor and the finger during attachment and detachment, resulting in decreased accuracy in force estimation. Therefore, in this study, we are focusing on mitigating the effects of attachment and detachment by incorporating the 3D shape of each individual's finger into the sensor design.

We are producing a tactile record and playback system to record and playback movements and manipulations of artisans and professionals even after their deaths. In this paper, we will report a three-axis tactile sensor, which is one of the tactile record and playback system’s key parts. The tactile sensor’s sensing principle is based on the light reflection occurring contact between a rubber sheet and an acrylic board. We introduce mathematical models to estimate three-axis force using brightness and centroid of bright area. As the results of loading tests, these models can estimate the three-axis force.

Micro screws with sizes M2.0 or smaller are widely used to assemble smartphones, tablets, and wearable devices. Due to the delicate nature of the assembled parts, the screws often require tightening torques less than 100 mNm with an accuracy in the range of ±5%. In recent years, high-performance electric drivers have been developed to meet these requirements. However, optimizing screw tightening conditions for micro screws of M2.0 or smaller remains challenging because the axial force during screw tightening cannot be directly measured using conventional methods. Consequently, the optimization process relies on time-consuming and costly trial-and-error methods. To address this issue, this study proposes a system that leverages IoT sensing technology and a machine-learning regression prediction model to accurately estimate the axial force during micro screw tightening. In the data collection phase, a system was designed to sense and capture multivariate time-series data, such as tightening torque, pushing force, and vibration of the electric screwdriver during screw fastening. Subsequently, a machine-learning regression prediction model was developed to accurately estimate the axial force of the screw based on the collected data. Experimental validation using M2 cross-recessed pan head screws demonstrated that the proposed system achieves axial force estimation accuracy within 10%. This study offers a potentially effective approach to optimize screw tightening conditions for micro screws, enabling efficient and reliable assembly processes in the production of electronic devices.

A steer-by-wire technology enables steering control of automobile by electrical signal without mechanical connection of tires and a steering wheel．This technology makes the characteristic of steering system change．Therefore，it can change relationship between steering angles and actual steer angles of tires according to the driver and driving conditions of vehicle．It has potentials to allow a driver rapidly and stably take action to avoid a collision if a driver must execute the emergency avoidance．The purpose of this research is to study possibilities of the improvement of the emergency avoidance performance by using a variable characteristics steering system（VCSS）as a steer-by-wire technology．Fuzzy control is used for the control of the VCSS．We conducted a simulation of an emergency lane change to avoid an obstacle on a road. We investigated vehicle motion，driver’s conditions and parameters of fuzzy control by comparing the vehicle having fuzzy control with the normal vehicle．The result of simulation shows the increase emergency avoidance performance by using the VCSS．

We have developed a pair of interactive devices for remote haptic detection and presentation using linear coil springs. Previously, we found that coil springs with a small spring constant provide accurate haptic reproduction but cannot reproduce large forces, and those with a large spring constant can reproduce large forces but shows inaccurate haptic reproduction. In this study, we proposed the use of conical springs to improve haptic reproduction since we hypothesized that the non-linear characteristics between deformation and the restoring force of the conical spring take advantage of linear coil springs with small and large spring constants. In the experiment, we applied loads to one of the devices and measured forces reproduced by another device to evaluate haptic reproduction. The result showed that the conical spring provided accurate haptic reproduction in a wider force range than the coil springs. Therefore, we concluded that nonlinear springs are suitable to improve haptic reproduction.

The purpose of this research is to develop a posture estimation method of nursing care givers using plantar pressure distribution. Because nursing care with improper posture can cause some diseases like a lower back pain, care givers had better be instructed about a proper posture of a nursing care in advance. Some researches of posture estimation of care givers have been conducted; nevertheless some sensors were required to be attached to the care givers’ bodies for the motion measurements. Conversely, our proposed posture estimation method does not require any sensors to be attached to the body; a pressure sensor sheet is only required to be settled under the nursing care area, and the posture is estimated from the plantar pressure distribution. That is, the positions of the feet and the height trend of the lower back are estimated from plantar pressure distribution by using a principal component analysis and a summation of the pressure distribution, respectively. In the experiments, a participant’s lifting up motion of a weight were conducted, and the postures of the participant were estimated by using the proposed method under the assumption that a care giver assisted a care recipient who moved from a bed to a wheelchair. As a result, the positions, the orientations of each foot and the height trend of the lower back of the care giver were estimated correctly.

Early detection and prevention of frailty are important for extending healthy life expectancy. Although the Japanese version of Cardiovascular health study (J-CHS) criteria are used to judge frailty, a more objective evaluation method has not been fully established. In this study, we focused on the Sit to Stand-five test (SS-5), which is often used in medical practice because it is a simple and easy way to evaluate lower limb function. We proposed a data processing method and an evaluation method to discriminate healthy, prefrailty and frailty using SS-5 data measured with a floor reaction force meter. We compare the proposed method with the J-CHS criteria to verify its usefulness.

The purpose of this study is to clarify human walking from the viewpoint of musculoskeletal mechanism. In the field of rehabilitation, Rocker function explains how human musculoskeletal system generates walking motion and efficient locomotion. Once these explanations were understood, we could reconstruct it by artificial materials. In this study, the human musculoskeletal leg was developed for simulating human walk. Muscles and tendons including biarticular structure were represented by springs and cables. The simulator walked by actuating several tendon cables with servomotor. In this paper, we introduced foot structure in the robotic simulator to represent the toe rocker in Rocker function. The weakened toe force resulted in shorting of step length.

Ultrasound (echo) guidance has been recognized as a standard practice for needle insertions to increase the safety and certainty of vascular access calculation in hemodialysis treatments. Nevertheless, its manual operation requires training and proficiency due to difficulties in adjusting the echo probe and needle to catch the correct identification of their spatial relationship from the two-dimensional cross-sectional images. This manuscript proposes an assistive technology for ultrasound-guided puncture by attaching a miniature inertial sensor unit (accelerometer and magnetometer) on both the needle and the probe. The method includes three-dimensional posture calculations to track needle orientation relative to the probe imaging direction. It also employed image processing to detect and track a needle coordinate of intersectional points in the ultrasound image utilizing optical flow algorithms. An optical motion capture system was used simultaneously for the validation study. The primary experiment in which clinical expertise punctured the skin-vessel model successfully showed the expected needle pass overlaying ultrasound image.

We have been studying two main methods for preventing patient falls and reducing the workload of medical staff at medical and nursing care facilities, using artificial intelligence and robotics technologies. One is a realtime fall risk assessment method of patients and their surrounding environment using deep learning. The other is a nursing assist method to present recommendations and nursing assist to medical staff for optimal interventions for patients using deep reinforcement learning to reduce the risk. In this paper, we summarize the results of our application study of representative nursing theories to deep reinforcement learning, and present the limitations of its application, problems, proposed solutions, and future prospects. We apply the following researchers’ theories: C. Roy，D. E. Johnson, D. E. Orem, V. A. Henderson, H. Peplau, I. Orlando, F. Abdellah, I. King, E. Wiedenbach, J. Travelbee, M. M. Leininger, M. Newman, M. H. Mishel, N. J. Pender, M. E. Rogers, R. R. Parse, J. Watson, K. Kolcaba, F. Nightingale, P. Benner, M. E. Levine, and develop it further. The result systematizes the positioning of each nursing theory, as well as gains new insights into the scope of the environment and the role of the agent, and by incorporating these findings, and this work advances the "harmonization of applicable environments”.

The purpose of this study is to make it safer and easier to get around using an automated wheelchair. In order to obtain highly accurate location information using RTK-GPS, a base station was set up at the university. We compared the errors of single positioning, RTK positioning using a distant base station, and RTK positioning using a base station installed in a facility in the university. As a result, the RTK-GPS system using the base station at the university had the highest accuracy. In the future, we plan to install RTK-GPS and LiDAR on a self-driving wheelchair to calculate the route to the destination.

The number of diabetes patients in the world is increasing, and it is estimated that the increase will continue in the future. Although shoes are prescribed for diabetics in the medical field, it is said that it is difficult to judge whether the shoes are suitable for the feet. The purpose of this research is to develop a system that can evaluate the shoe fit of the entire foot in real time from the shear force measurement by vibration. Using a piezoelectric wire sensor, we measured and analyzed the vibration inside the shoe and verified whether it is possible to detect the difference in shear force generated during walking. In addition, we verified whether it is possible to improve the classification accuracy by machine learning by setting different numbers of divisions for the first half and the second half of the frequency band in which the frequency range of the vibration generated when the shear force is generated.

A tool to control the bending of a laser fiber protruding from the endoscope tip using shape memory alloy(SMA) wire that shrinks when heated by Joule heat was developed. In the conventional bending device, the SMA wire that is driven during bending contacts the outer tube that covers the device and overheats the surface of the device. For the prevention of surface overheating, a link part was added to the center of the device that restricts the movement of the driven SMA wires. The bending device with a center link was fabricated and measured the bending angle and surface temperature. The device with a center link realized a larger bending angle and lower surface temperature comparing the conventional structure.

In this study, we develop a human-following robot for the purpose of life support in indoor living spaces. The robot uses thermal image data obtained from a temperature sensor to distinguish between human and object. In living spaces, there are high heat sources that exceed the temperature of the human body, such as heaters, and contact of the robot with these high heat sources may lead to accidents such as fires. Therefore, it is necessary for the robot to avoid high heat sources while following the human. In this study, we propose a method in which a robot equipped with distance sensors, a geomagnetic sensor, and infrared heat sensors can distinguish between room temperature and heat sources and avoid high heat sources by identifying high and low heat sources at the same time.

The main cause of accidents occurring in works is that the unexpected high loads applied to the body during several motion. Visualization and quantification of the load acting on body have potentials to prevent the accidents before happens and to develop assistive technologies. Surface electromyography is often used to evaluate the muscle activities in the rehabilitation and sports fields. However, the electric signals detected through the electrodes attached at the human body surface are susceptible to the noise with the presence of sweat and body hair. Therefore, the prior skin treatment is required to avoid the bad electrical contacts. Recently, force myography (FMG) has been proposed as a simple measurement method by using the detection of contact force between the sensor and the body. In this study, an FMG-based pushing device which measures the reaction force relating the change of body hardness was fabricated. And muscle activities in the upper limb were measured during walk with holding a heavy load. As a result, the effects to lift the weight and to up the stair showed in the muscle activity evaluated in FMG measurements.

In laparoscopic surgery, it is difficult to treat deep inside the body because of the rigid and straight surgical tools. Thus, we have proposed a surgical system using cassette-type flexible surgical tools and a guide tube that can change the rigidity of the body for securing the path to the affected area. So far, we have developed the components of the cassette-type flexible surgical system: cassette-type flexible surgical tools, the rigidity changeable guide tube, a forceps actuating device, and a forceps exchanging device. However, the forceps exchanging device is still large and heavy. Therefore, in order to reduce the size and the weight of the forceps exchanging device, we proposed and developed a forceps exchanging device with a rotating cassette holder. Moreover, we evaluated the selecting forceps function. As a result, we observed that the device can select each cassette of the five cassettes placed inside the exchanging device in around three seconds.

Myoelectric control system is a fundamental component of modern prosthetic devices, which uses the myoelectric signals from an individual’s muscles to control the prosthetic’s movements. Currently, surface electromyography (EMG) is the most popular myoelectric control used in prosthetics to detect electrical signals generated by muscles. (MMG) is a less known technique that involves detecting and interpreting mechanical vibrations generated by muscle contractions. This technique requires no electrical connection to muscle tissues, thus making them more resistant to noise interference and is also a cheaper alternative. This study intends to investigate an alternative method to classify hand gestures for prosthetic control by simultaneous acquisition of EMG and MMG signals. The sensors used are two Myoware Muscle sensors to record EMG signals and two MPU6050 accelerometers to record MMG signals. The hand gestures measured are gripping, extension, flexion, supination, and pronation. Pattern recognition using long short-term memory (LSTM) method is used to classify hand gestures.

With the resulting overall accuracy of 86.8%, this study confirms that the simultaneous acquisition of EMG and MMG signals has the potential for anthropomorphic prosthetic control.

The goal of preventing 50% of traffic accidents worldwide has been set. Japan is also implementing traffic safety initiatives. Elderly people aged 65 or older account for 57.7% of traffic fatalities in Japan. One of the reasons for this is that their reaction time is slower than that of younger people. Therefore, it is necessary to increase the reaction speed. Surface EMG, a type of biological signal, is a signal that appears before the start of a movement. In this study, we used machine learning to estimate the intention to move.

As technology progresses and robots become more common and accessible, we need better and more intuitive ways of controlling them. Haptics provide the sensory feedback needed to control robots with the same degree of confidence as we control our own bodies. An experimental master glove and slave robotic hand pair is used to test this approach. This is achieved using a motor that acts as both a force actuator and a position sensor, immersing the user in the slave robotic hand’s environment, allowing a much more natural movement and force output on the slave robotic hand.

In stroke upper limb rehabilitation, it is important to recover reach-to-grasp tasks, which is the core of upper limb function. Proper recognition of the spatial location of an object is necessary for reach-to-grasp tasks. Therefore, we developed the system that can evaluate the stroke-specific impairments based on time measurement in reach-to-grasp tasks within a specific area. Before the clinical application, we confirmed that the system's synchronization and measurement capabilities are accurate enough to perform measurements on the order of 1ms. The results indicated the system performed with good agreement for a specified spec. This paper describes the details of the system and the evaluation experiments.

Currently, the number of elderly people is increasing in Japan. Among people with disabilities, the most common type is limb impairment. It is expected that the demand for wheelchairs will further rise in the future. In our previous research, a mechanism was developed to increase the radius of the wheels only when climbing stairs. To evaluate the performance of the developed wheels, experiments were conducted where stairs were pulled and climbed. However, since the pulling was done manually, the pulling force varied with each trial. In this study, an electric pulling device was created, and experiments were conducted to climb stairs with a constant pulling force. As a result, quantitative evaluation was possible by using a motor. It was also found that the contact position between the step and part of a wheel (named foot) affected the performance of the developed wheel to climb the stair.

For safe driving in an environment where autonomous vehicles and other vehicles coexist, we have constructed a "safe behavior design diagram" that examines the risk reduction of unsafe human behavior. In the present study, the ultimate goal is to use autonomous vehicles safely, and based on the causal loop diagram, we hypothetically dealt with the transportation of goods by autonomous vehicles at the transportation site. The work was analyzed by "Antecedent - Behavior – Consequence (the A-B-C model)" by the behavior analysis method. After that, we extracted possible problems from the analysis, performed risk assessment for use in the area of machine safety, and tried to distribute the risks appropriately by dividing them into each step.

When a traffic jam occurs on the main line road, many vehicles, such as compact vehicles or light motor vehicles, etc., which have a small turning radius, try to detour to nearby side roads where they can run smoothly. In this study, authors applied the function of ETC2.0 and measured the velocity change and traffic volume in each section by dividing the road into small sections of about several meters. Then, authors analyzed traffic jam conditions of main line road and the occurrence of detour traffic of vehicles entering residential areal roads to avoid traffic jam, and obtained the changes in the traffic jam conditions and the traffic volume on residential areal roads depending on the time zone.

Logistics is considered to be the delivery of needed goods to the needed place (or person) at the needed time． In this study, we will design an automated logistics system using Work Design, a design thinking method． For this purpose, "logistics is to deliver needed goods to where (or to whom) they are needed, when they are needed, in as automated a manner as possible" is added before the last part of the definition above． In Work Design, logistics operations are designed in the order of time． In addition, the order is always repeated from the whole to the design of the parts． There are several other features, but it is thought that designing by Work Design facilitates the emergence of automation ideas．

The number of traffic accidents in Japan tends to decrease because of deployment of advanced driver assistance systems and active safety systems. However, accidents involving pedestrians and bicyclists account for a large number of fatalities, and further countermeasures to avoid those accidents are necessary. There are several cases where collisions with pedestrians and bicyclists cannot be avoided with the current specification of existing systems. In order to reduce the number of traffic accidents while walking or riding a bicycle, further improvement of driver assistance systems is required. This paper proposes a potential risk index called accumulated risk to predict theoretical avoidance behavior based on the relationship between the acceleration amplitude for planned trajectory and the collision risk. In order to obtain the reference acceleration for each driving scene, risk avoidance behaviors were formulated according to the predetermined patterns of acceleration in both longitudinal and lateral directions. The formulation of the avoidance behavior enables the calculation of acceleration command values to coordinate with other ADAS functions.

Personal mobility vehicles (PMV) attract our attention as a countermeasure to the social problems caused by the spread of automobiles and the serious aging society. Since PMVs have smaller widths than that of ordinary vehicles, it allows drivers to drive with flexibility of a running position within the lane. It is, therefore, necessary to investigate which running position in the lane is safe to drive. In addition, the installation of some autonomous driving systems is advanced for ordinary vehicles. It is likely that PMVs will also be equipped with driver assist systems or autonomous driving systems. In this paper, we create a model of an automatic collision avoidance system with both of an automatic braking and an automatic steering for a PMV, and simulated obstacle avoidance situations in three different running positions in the lane (center, left side and right side). As a result, the investigation of the avoidance success rate shows the running position affects the avoidance characteristics of the PMV.

This research aims to develop a new water surface cleaning ship that can efficiently collect floating matter of various sizes and liquids floating matter such as oil on the water surface not only in forward and backward motion but also stationary state. The proposed water surface cleaning ship is not a mechanical type, but can capture floating matter by using the flow generated by the water level difference inside the ship by means of a control free movable weir the results of the water level difference measurement test showed that the higher the weir height of the movable weir, the greater the water level difference, regardless of the length of the net of the floating matter recovery device installed inside the scaled model ship.

This study aims to develop an advanced driver assistance system (ADAS) to avoid collision with an oncoming darting-out vehicle from a blind area when turning right at intersections. As commercialized technology, ADAS activates Autonomous Emergency Braking to avoid collisions when on-board sensors detect oncoming vehicles during right turns. However, the system may not avoid a collision with a target that suddenly darts out from a blind area. Connected vehicles in coordination with infrastructure have been commercialized as ADAS technologies for this purpose, but they cannot provide assistance unless installing communication equipment for both the vehicle and infrastructure. Therefore, this paper proposed an ADAS that uses on-board sensors without depending on infrastructure facilities to control the vehicle speed to avoid collisions. This system calculates the dangerous speed criterion of the ego vehicle that has a high risk of collision when a darting out occurs and provides speed assistance to escape from the dangerous speed region. Simulation results showed that the proposed system effectively reduces the possibility of collisions.