We have developed a design insight technology for design documents in order to reduce overlooking of failure modes when making design documents. This technology extracts parts from design document and shows a list of parts and a fault tree with score for a part selected by user. The fault tree gives awareness of failure modes for a user. The score is calculated considering field and product, operating environment, impact based on number of co-occurrence between field and product, operating environment, impact and failure mode. A user can prevent failure modes efficiently by considering from higher scores. As a result of verification, we confirmed that the score changes according to the field and is calculated properly.

The Frictional vibration called a stick-slip occurs on the sliding surface. A familiar example of this vibration phenomenon is the chattering vibration of machine tool. When the stick-slip occurs, it leads to reduce a safety, a life expectancy and a work precision. This phenomenon has been studied for a long time, but the mechanism is not fully understood because the sliding friction behavior is caused by various factors. Therefore, it is necessary to have an analysis method that can represent the state of objects and contact surfaces. In this study, the focus was on rubber blocks, which are prone to stick-slip due to the large deformation of the object. A friction model and a deformation model in the DEM used for analysis and investigation. And, a model for analyzing the stick-slip behavior of rubber blocks is proposed. The results showed that the period and horizontal amplitude of stick-slip tended to be closer to the experimental results when the element size was reduced.

The equations of motion for manipulators are important in simulations. In the set of equations of motion for a manipulator the structural parameters of the manipulator and the inertia parameters of the links appear. The structural parameters are obtained from the design data, while the inertia parameters are obtained by estimating the values from the input torque and output data. When all the inertia parameter values of the equations of motion are obtained, some combinations of the obtained parameter values may become physically impossible. In this study, we investigate the physical impossibility of sets of base-parameter values for a 3d,o,f manipulator.

Vacuum molding process is suitable for large thermoplastic parts production. The molding time of this process is relatively long, so the creep deformation and stress relaxation are occurred simultaneously. Since forming conditions affect dimensional accuracy, trial and error are generally necessary for setting the molding conditions.

The purpose of this study is to develop the forming simulation technology considering creep deformation and stress relaxation phenomena in vacuum forming, and the applicability of the “Generalized Maxwell Model” to thermoplastics, which are viscoelastic materials, was verified. Concretely, on dimensional change of element test loaded and unloaded with constant stress, structural analysis results incorporating “Generalized Maxwell Model” was compared with creep test results to evaluate validity of analysis.

Two methods, power injection method (PIM) and approximate power injection method (APIM), are analytically compared in input-power identified of the simplest two subsystem SEA model. From the formula determined by APIM, it is recognized that it is happened to identify the power input into the subsystem without actual input power. Then we discuss and propose an improved SEA model construction method based on APIM. The improved APIM is numerically applied to evaluate the loss factors and the power inputs of a simply supported beam considered by two SEA subsystems. Also the improved APIM is experimentally applied to identify the power input of the automotive tire with two SEA subsystems and it is confirmed that the identified power inputs are more accurate.

Model-Based Development(MBD) is attracting attention as tools for further efficiency and improvement of designing quality in these days while product is highly functionalized and diversified. MBD provide improvement of designing quality at an early stage of product development through "front loading" by using simple physical model. This study performed MBD for fuser-nip for laser printer and achieved more detailed design in virtually than before by using simple simulation. The simulation was based on not only heat transfer but also viscoelasticity of tonner. This paper shows a series of details, construction of model, development of fuser-nip and verification.

Titanium alloys have excellent specific strength, corrosion resistance, and biocompatibility. Because of these characteristics, they are widely used in aircraft, chemical plants, sports equipment, and medical treatment. Beta-type titanium alloy has a body-centered cubic lattice structure at room temperature, which means that there are many slip planes in the crystal and plastic working is easy. However, by heat treatment, this alloy exhibits the greatest hardness among titanium alloys. This alloy has about twice the strength and modulus of elasticity of austenitic stainless steel. In the present study, the effect of stress shot peening on the fatigue strength of β-titanium alloy was investigated. Stress shot peening can further increase the residual compressive stress, thus reducing fatigue crack initiation. This peening produces large compressive stresses throughout the compression layer. The test material was a commercially available beta-type titanium alloy. The stress applied was less than proof strength. Shot peening was done with an air-type machine. The microshot media was made of cast steel with an average diameter of 0.1 mm. Projection pressure and time were 0.6 MPa and 60 seconds, respectively. The fatigue strength of the stress-treated workpiece was about 1.9 times greater than that of the untreated workpiece. Stress shot peening treatment was effective in improving the fatigue strength of beta-type titanium alloy.

Austenitic stainless steels are resistant to rust because of their ability to form a very thin protective film. In addition to corrosion resistance, it has excellent mechanical properties, workability, and heat resistance. Stress corrosion cracking and intercrystalline corrosion can be a problem in austenitic stainless steel products. In order to prevent stress corrosion cracking, solution heat treatment is applied to remove residual stress after plastic forming. After that, annealing at about 400°C is generally carried out. However, some of the various parts are still machined or polished, and the material surface has an affected layer. The effect of heat treatment on materials with an affected layer has not been well investigated. Furthermore, the fatigue strength of heat-treated materials has not been well investigated. In the present study, the effect of shot peening on fatigue strength of low-temperature heat treated stainless steel was investigated. The test materials were four commercially available austenitic stainless steels. The fatigue tests were performed by rotating bending fatigue testing machine. The low-temperature heat treatment temperature was 400°C. Microshot peening was done by an air-type machine. The media was made of cast steel with an average diameter of 0.1 mm. Air pressure and peening time were 0.6 MPa and 60 seconds, respectively. The fatigue strength of the low-temperature heat-treated workpieces increased by 1.3 to 2.7 times compared to that of the untreated workpieces. The fatigue strength of the shot-peened treated workpieces increased by a factor of 1.9 to 3.4 compared to that of the untreated workpieces.

Mg alloys are increasingly being used in the field of transportation equipment, such as automobiles and aircraft, where energy saving is being promoted. However, the problem with Mg alloys is that they have lower corrosion and wear resistance than steel materials. Therefore, a lot of research on alloy development and surface modification has been attempted. In this study, we attempted to modify the surface of magnesium alloy by using shot peening. The processing machine was a centrifugal type, and the projectile was made of cast steel with an average diameter of 1 mm. The shot speed was 60m/s and the peening time was 20s. The ceramic particles are alumina and zirconia, with a particle size of 0.1 to 0.2 mm. The resin is a sheet of polyimide, with a thickness of 0.025 to 0.05 mm. Corrosion tests were conducted by immersion in sodium chloride solution. The bonded dissimilar metal foils were tightly bonded to the base material. No voids or cracks were observed at the bonding interface. As a result of bonding resin sheets containing ceramic particles, the corrosion and wear resistance of the base material was improved.

Magnesium alloys are characterized by their light weight, high specific strength and high specific rigidity. Therefore, it is expected to be an effective material for energy saving in the field of transportation equipment. However, the corrosion and wear resistance of magnesium alloys is much lower than those of steel materials. For this reason, surface modification is performed using surface treatment techniques such as vapor deposition and plating. In the present study, dissimilar materials were joined using particle impact such as shot peening to modify the surface of magnesium alloy. The test material was a commercially available magnesium alloy, and the dissimilar metal sheets are pure aluminum, pure copper, pure nickel, pure iron, and stainless steel. The thickness of the various thin sheets ranges from 0.02 to 0.06 mm. Shot peening was done with mechanical equipment. The projectiles had an average diameter of 1 mm and were made of cast steel. The processing speed was 60 m/s, and the processing time was 20 s. The joinabilities of the thin sheets were evaluated by cross-sectional observation and bending tests. When the processing temperature was 300 oC or higher, various types of thin sheets could be bonded to the base material. There were no voids or cracks at the bonding interface. After the bending test, the tightly bonded metal sheet broke along with the base material.

In this study, the effects of the surface treatment condition on the bonding strength in the ABS resin directly formed on the aluminum alloy were investigated. The ABS resin filament was deposited as a single line on the aluminum alloy by a fused deposition modeling method. The shear bonding tests of the ABS-resin bonded on the aluminum-alloy were performed. As the surface treatment condition of the aluminum alloy, the anodic oxidation and laser roughening process were selected. It was found from the experimental res that the bonding strength of the joints was increased by the anodic oxidation treatment with the sufficient treatment time since the growth of the micropore structure with the treatment time. The enlargement process of the nanohole on the aluminum surface created by the anodizing process was also increased the bonding strength. However, if the treatment time became longer than the critical time, the bonding strength was decreased due to the excessive growth of the structure reduced the wettability of the resin to the aluminum alloy. The deterioration of wettability by the anodizing treatment could be decreased by applying the vibration of the aluminum alloy during the deposition of the resin filament. The laser roughening process achieved higher bonding strength due to the expanded anchoring effect. In this study, no significant difference was observed by the laser roughness process condition.

Drawing is a type of press work that uses punches and dies to create a seamless container-shaped product from a single flat plate. This is used not only for containers, but also for industrial products and machine parts because it can be used to process a variety of cylindrical shapes and square cylinders. In the present study, an attempt was made to form a pure titanium corrugated cup with a structure similar to a corrugated paper cross-section for the forming of lightweight and high-strength functional cups. The test material was commercially available pure titanium, ultralow-carbon steel, and stainless steel. The shape of the blank was a disc with a diameter of 80 to 100 mm. The deep drawing machine used was a hydraulic machine. To form the cup into a corrugated shape, the roller ball die was used. The clad cup was formed by a composite die with a structure consisting of the roller ball die sandwiched between two regular dies. The formability of the drawn cups was evaluated by measuring thickness strain and compressive strength. Load-displacement measurements were also taken to investigate the forming transition. No fracture occurred in the forming of the corrugated cup. The thickness reduction at the bottom of the cup was less than 10 %.

Pure titanium has high specific strength and high corrosion resistance, and is also lighter than stainless steel in corrosion resistance. It has been widely used together with titanium alloys for a long time in chemical plants and heat exchangers. Pure titanium, which is widely used as a product, is JIS type 2. It is expected that JIS type 2, which has high strength while maintaining purity, will be used. In general, cold working and heat treatment methods are used to increase the strength. In the present study, the material properties of pure titanium were improved by a combination of cold rolling and heat treatment. Mechanical properties were examined by tensile testing. The test material was a commercially available pure titanium JIS 2 sheet. The initial thickness of the sheet was 1 to 5 mm. The thickness of the tensile test workpiece was 0.4 to 0.7 mm. The rolling reduction ranged from 30 to 92 %. An increase in tensile strength was obtained in the heat treatment from 400 to 500 oC while maintaining the ductility. In particular, the tensile strength was greatly improved under the condition of 70 % of rolling reduction and heat treatment temperature of 400 oC. The effect of strain rate on mechanical properties showed little effect.

Welding is used in many structures. Since heat is applied locally to the material in welding, tensile residual stress is generated near the bead, causes the stress corrosion cracking in stainless steel. We have developed welding residual stress reduction for structural rolled steel using ultrasonic vibration load. In this study, we apply this method for stainless steel SUS304 block material, and measure welding residual stress. Comparing the results of residual stress measurement between with and without ultrasonic vibration load, effectiveness of this method on stainless steel is investigated. As a result, this method can reduce welding residual stress in stainless steel. The reduction rate in case with acceleration amplitude of 5000m/s² is greater than one in case with acceleration amplitude of 2000m/s². From the t-test results, significant differences of welding residual stress reduction between them were confirmed. In case of ultrasonic vibration load with acceleration amplitude of 2000m/s², welding residual stress was reduced 16% compared to no load. In case of 5000m/s², it was reduced to 30%.

In this study, TZM alloy and Titanium Carbide (TiC) and Titanium Nitride (TiN) added TZM alloy sintered bodies were fabricated by spark plasma sintering. Density, Vickers hardness and average grain size of sintered bodies were evaluated. Vickers hardness values increased with an increase in containing of additives in TZM. TiC added TZM sintered bodies had higher hardness compared to TiN added TZM sintered body. This difference could be due to difference in hardness of the additives. Densities decreased with increasing in containing of additives, because additive has lower density compared to TZM alloy matrix. Precipitates with size approximately 1μm were observed mainly at grain boundary in TZM-10vol%TiC sintered body. This size of precipitate was corresponding to primary particle size of TiC. It was speculated that TiC additives were mainly dispersed along grain boundaries. Value of Vickers hardness predicted with compound rule compared to that obtained with experiment. Experimental values were higher than calculated values. TiC and TiN added TZM alloy sintered bodies were strengthened by dispersion strengthening mechanism of additives but not only mixing of hard particles.

In general, removal machining cannot fabricate a hole on the inside wall of another hole. This means that removal machining has the limit on the degree of freedom of its fabricatable hole shape. To overcome the limit, our laboratory aims at the establishment of the method to fabricate such holes. Its previous studies developed the devices which can fabricate the holes by using with a die-sinking electrical discharge machine (EDM). However, those studies dealt with the improvements of the length, diameter, and straightness of the fabricated hole, and did not deal with the enhancement of the surface roughness of the fabricated hole. So, this study has developed the method to enhance the surface roughness by realizing the finish machining of the inside wall of the hole. The experiment result proves that the developed method has the abilities to finish the inside wall of the hole and to enhance the surface roughness of the hole.

Our laboratory aims at the development of curved hole electrical discharge machining (EDM) system by means of a microrobot. The microrobot is composed of autonomous EDM unit, direction determining unit, and in-pipe self-mobile unit. The previous studies have developed these units independently. In these studies, the autonomous EDM unit has developed which uses shape memory alloy (SMA) as its actuator and the in-pipe self-mobile unit has developed which employs hydraulic cylinders as its actuator. So, this study has developed the device combining these two units and the control system which makes the device drill a hole autonomously. The experiment result shows that the developed device and system have the ability to autonomously machine a hole.

There are many irrigation canals in the whole country. In Okayama, accidents to fall to the irrigation canal occur frequently from around 2013. The fall accident of 1,143 cases occurs from 2013 through 2015, and 79 cases of them result in death. 50 years or older accounts for 90% （70 years or older is near 70%）, and outbreak in the bright time of 6:00 p.m. goes up the age that They were in an accident to 80% from 6:00 a.m. It entails a large amount of cost because a lot of irrigation canals are greatly made from other prefectures in Okayama where agriculture was prosperous if I touch a fence in all waterways. In addition, it is very difficult to take measures in a lump whether a prefectural road is along it whether the management of the irrigation canal is along a city street or whether it is along a private road because the jurisdiction varies according to places. In addition, it is difficult to take perfect measures because it may be difficult to irrigation canals that are used as agricultural water in some places. Therefore it is necessary to be aware of hazards and consider the safety of people in places where fences are difficult to install. In this study, the authors propose a system that uses light and sound to warn of danger and prevent people from falling when multiple infrared motion sensors are used to detect danger.

Currently, elderly people account for more than 25% of the population in Japan, becoming a super aged society. The problem of Locomotive syndrome accompanying aging is a problem, and its prevention and improvement is considered an urgent issue. In order to realize the extension of healthy life without deterioration of quality of life (QOL), this research focuses on Sarcopenia and tries to build countermeasures for it. Sarcopenia is that muscle weakness of the whole body occurs. When it attenuates the muscles of the lower limbs, walking becomes difficult. In order to prevent this, in this research, we aim at rehabilitation by walking using a treadmill, aiming at strengthening the muscles of the lower limbs and restoring gait performance. Here, we focused on Soleus muscle and developed a self-propelled “Load-controlled treadmill” aimed at enhancing its muscles. Since the walking posture and speed are different between the Load-controlled treadmill and a conventional treadmill, walking experiments were carried out to compare these treadmills. In addition, we investigated the difference in gait depending on load patterns of the Load-controlled treadmill. It was evaluated by the change in the muscle activity amount of the lower limb muscle, the joint angle and the joint moment. As a result, it was found that the Load-controlled treadmill induces plantar flexion of ankle joint and activates the muscle activity of the lower limbs. Also, fatigue was reduced in some load patterns.

For rehabilitation or locomotor training purpose, treadmill training is widely used since it only requires a small space and has several benefits such as improved heart health, increased muscle strength, etc. However, since treadmill training is repetitive, it can become tedious and discourage users from training regularly. In this paper, we propose an immersive walking training system that uses virtual reality technologies and a treadmill that is automatically controlled by the user’s walking speed. This system is expected to create more engagement, and enable users to enjoy the training, thus increasing the training frequency as a result. Actual environments recorded as a 360° video can also be used, enabling people who are house-ridden to enjoy an outdoor experience.

The aim of this study is to increase the easiness to grip the T-shaped walking stick. For this purpose, we developed a measurement system for cane walk, and using this system we measured strain on the cane, and EMG during cane walk, as well as sensory evaluation after walking. In the experiment, seven healthy adults were set to walk using 5 types of grips during walking, with a stride of 30 cm and a tempo of 60 BPM. As a result of the experiment, the strain voltage acting on the grip of the horizontal oval-shape grip was significantly smaller than that of the round-shape and vertical oval-shape. In addition, the horizontal oval-shape had the best evaluation in the sensory evaluation, such as the SD method. This shape reduces the muscle burden of the gripping motion, and even if the wand is held for a long time, it does not get tired easily. We believe that we were able to obtain new knowledge that will lead to development of a walking stick that is less tiring.

It is important for the patients with neurological difficulties such as ALS and SMA to improve their quality of life. We focused on the sport called Boccia that can be played even by people with relatively severe disabilities, and the ultimate goal is to construct the system for patients with intractable diseases. In this research, we developed the Boccia play support system which can allow the disabled player to determine both the direction of the lamp and the launch height of the ball, and to launch the ball using movable body part, on the assumption that it can be remotely controlled. Here we paid attention to the slight movement of fingertip. The skin surface on the fingertips was observed with USB camera, and the amount of movement of feature points on the skin surface was derived by using optical flow, and a non-contact interface based on skin strain detection was developed.

In industrialized countries, aiming at assisting heavy labor workers, assistive robots for heavy labor have been actively developed to reduce occupational accidents considering not only the spine load but also the finger jamming. In this study, we developed a wearable work support system to reduce the load on the waist and the stuffing of fingers in construction work. A furnace construction work was analyzed mathematically to derive the load applied to the body and an effective method was developed to reduce the load. Finally, we developed a wearable work support system with a vacuum gripper that reduced the load on the lumbar region through the pelvis instead of holding heavy objects with the lumbar muscles and prevented finger jamming during furnace construction work. As a result, the demonstration experiment confirmed the reduction in the amount of muscle exertion with the developed system and the prevention of finger stuffing by improving the brick gripping method using the vacuum gripper.

In recent years, many driver assistance devices evolve and become widespread, the information has, however, become more complex. There is a danger that drivers may not be able to process the information and may lead to accidents. In order to solve this problem, we are developing an electronic manual, defined as an "active manual," which detects skills and comprehension levels of the driver and actively gives adequate information to the driver. In the experiment, a driving simulator is used to observe the behavior of subjects driving and detect their various levels of understanding. As a result, it is found that the real-time information of the active manual is effective for drivers with low skill and comprehension level.

In our previous study, we developed a fingernail color sensor capable of identifying the magnitude and direction of the force acted on fingers. The prototype is equipped with a small CMOS camera and green LEDs, and collected a dataset of vertical force and shear force acting on the index finger and fingernail color images using an electronic balance and a load cell. The dataset was trained by CNN to estimate forces from the images. In this study we produced individual CNN models, which learned individual fingernail color change data. RMSEs for nine human subjects were 0.12, 0.12 and 0.29 [N] for F_x, F_y and F_z, respectively. RMSEs of F_x and F_y are significantly smaller than those of previous studies.

The simple measurement of body water content has the potential to be applied to new prevention methods and risk assessment for various diseases, and also important as a clinical indicator. In this study, we conducted a physical model experiment using water balloons with a non-contact water content measurement device using microwaves. We measured the phase difference of water accumulation in the physical model. The physical model was considered as a sphere, and the change in the diameter was theoretically calculated. The theoretical value of the diameter and the measured phase difference were compared and evaluated. It was suggested that the volume change in the physical model could be measured by the phase difference, which would make it possible to measure the water content.

Spastic gait caused by central nervous system diseases in infants is not easy to distinguish from internal rotation gait. Spastic gait is a problem with commands from the brain that causes gait abnormalities, and internal rotation gait is a problem with joints that causes internal crotch gait. These diseases are diagnosed by walking posture, rhythm, stride length. When making a definitive diagnosis of CNS disease in early childhood, MRI is commonly used. Therefore, MRI is invasive and dangerous. A safer and simpler method for distinguishing the cause of gait abnormalities is desired. In this study, as the first step to develop the method that distinguishes spastic gait from internal rotation gait, the characteristics of the internal rotation gait was examined. During the experiment, the two types of gait were measured using an optical motion capture system; normal gait and internal rotation gait performed by a healthy participant. Singular value decomposition was applied to an observation matrix consisting of the reflective marker coordinates of the lower limb obtained in the experiment. The results demonstrated that the movements of arbitrary markers were coordinated during internal rotation gait.

For patients with muscular dystrophy, a motor dysfunction, who have difficulty operating an electric wheelchair with joysticks, a simplified one-input device is used. However, avoiding obstacles can be time-consuming. In this study, we analyzed the motor functions of the thumb of a patient with severe muscular dystrophy and identified the operations that did not cause physical fatigue. Then, we developed an operation support system to continuously operate. Finally, we conducted experiments comparing the proposed system with the conventional system and verified the effectiveness of the proposed system based on the steering accuracy of the electric wheelchair and the task completion time.

The incidence of back pain in the health care industry is the highest among all industries, and the number of cases of back pain is increasing every year. To prevent the back pain during the nursing care, it is very important to understand and keep the correct posture and movement in the care. However, it is difficult to keep the correct posture and movement in a real situation, even though there are opportunities to learn nursing care movements that are less likely to cause back pain. In response to these issues in nursing care movements, visualization of ideal nursing care movements that do not cause back pain and quantification of the burden on the lumbar region have been clarified one of measures to prevent back pain. However, although it is possible to grasp the tendency of the ideal posture that is less likely to cause back pain, the accuracy of this method has been considered problematic. Therefore, we improved the accuracy of skeletal estimation in visualization by using depth information for photographing nursing care movements and quantified the burden on the lower back mechanically by deriving the moment applied to the lower back.

To reduce fall risks for the patients in medical and nursing care facilities, I have proposed and developed a nursing assistant system applying artificial intelligence and robotics technologies that predict incidents and assist inventions by automatic risk-assessing the condition of patients and environments. However, it is very difficult to make judgments and actions properly even if multi-professional cooperation level recognition can be done depending on the situation for medical staff. Therefore, in this paper, we introduce a method of applying nursing theory using deep reinforcement learning to the recognition / judgment / control model mounted on the autonomous mobile nursing assist robots.

The sense of touch plays an important role in the evaluation of products. However, there are individual differences in the degree of evaluation and discrimination ability, making tactile sensation difficult to share with others. Therefore, it is necessary to elucidate the factors of individual differences in tactile sensation and to enable quantitative evaluation of tactile sensation. In this study, we focused on the mechanical properties of the stratum corneum of the fingertip as a factor of individual differences in tactile sensation. We constructed a method to measure the thickness and Young's modulus of the stratum corneum. By using optical coherence tomography, we obtain the tomographic images of the fingertip during contact with an object. By preliminary experiments, it was confirmed that the constructed measurement system can be used to measure fingertip skin deformation in response to pressing force and friction force.

In this study, we investigated the conditions under which stones are efficiently scraped off by friction by vibrating the scalpel at high speed with ultrasonic waves. In the previous research, the phenomenon of cutting stagnation in a stepped manner was observed, but only one condition was performed. If there is a difference in the stagnation phenomenon depending on the cutting conditions, it is thought that a treatment procedure that can cut more quickly can be found. I studied. As a result, it was found that the larger the pressing force, the higher the cutting rate, the larger the final cutting amount, and the larger the pressing force, the shorter the time when the scalpel was about to stop.

The tactile feeling is important in product development because it influences the consumer's willingness to buy. In particular, the tactile sensation depends on the surface properties such as the spacing and height of the texture convexity, and texture design has attractive attention. Decorative treatments are one of texture design method. In some cases, due to performance or physical limitations of the material itself, material may be used to represent the target material and the surface is decorated for imitation. In this study, we focused on the tactile imitation of wood by resin print, which is one of the decorative treatments, and clarified the relationship between the surface properties of resin-printed samples and human perception of material texture. We conducted sensory evaluation experiments on resin-printed samples to clarify types of sensations that contributes to a sense of wood and investigated the relationship between the amount of sensation and the sensor output. As a result, it was suggested that roughness contributes to the sense of wood, and that the sense of wood may increase as the roughness increases.

In a system that records and distributes images using a wearable camera during surgery in a medical setting, we investigated a system using sensor feedback and a sensor-less system to prevent VIMS by correcting shaking of the surgical field images and shifts in the angle of view. In this study, we compared the accuracy and practicality of two sensor feedback systems: one that uses 6-axis acceleration and angular rate sensors to measure the position and posture and correct the image, and the other that uses image matching processing such as SIFT to correct the image without using sensors.

The purpose of this study is to decode complex sounds given as auditory stimulus from the brain image acquired by fMRI. In this report, we propose a system for estimating complex sounds using 3DCNN, one of the deep learning methods. In the proposed system, to estimate complex sounds composed of multiple pitches, a two-output 3DCNN model that can identify only sound of specific pitch was created as a deep learning template for each pitches. And the scale of auditory stimulus is estimated combining the sound of specific pitch detected by the deep learning template for each pitches. As verification, 20 kinds of compound sounds with 3 sounds selected from 6 different pitches were estimated using the proposed system. The deep learning templates were created using brain image of one subject for every six pitches. As the result, the maximum estimation rate was 45.00%. And the average estimation rate was 34.72%. Although the estimation rate of each deep learning template needs to be improved, we think that the structure of superimposing a deep learning template on each pitches is useful.

Toyota Motor Corporation and Woven Core, Inc. have continued research and development of automated driving technology and commercialized Advanced Drive using intelligence technologies. This system uses a high-definition map for highly precise self-localization and has a large-scale redundant architecture for recognition, control computation, actuators, communications and power supplies, to enable safe, reliable, and efficient driving on highways. The system also includes digital data uploading and downloading capabilities wirelessly over-the-air (OTA) in order to provide customers new software features and upgraded performance as well as the latest map updates.

In the risk potential driver model used to generate path control targets for autonomous vehicles, traffic elements such as obstacles and traffic rules are represented by risk potential, and safe routes are determined by synthesizing them. Therefore, it is possible to calculate the control target path adapted to the individual by quantifying the feel of risk that the human driver perceives and using it as a parameter of the driver model. This paper describes a method for estimating risk potential parameters using human driving maneuver. First, the travel path was functionally approximated to create a smooth trajectory similar to a skilled driver, and the steering operation to obtain this trajectory was calculated. Next, using the proposed method, the risk potential parameters were estimated using the nonlinear least squares method. It was confirmed that the driver model to which the estimation results were applied can reproduce the driving behavior of a human driver with high accuracy.

In this study, we analyzed the human sense of speed with respect to changes in the field of view by presenting a partially hidden expressway driving image that was reproduced through virtual reality. The obtained results indicate that even if the visual field range is constant, there are varying tendencies of change that depend on conditions such as the driving speed and road environment. Furthermore, such tendencies are supported by human visual field characteristics and distance perception mechanisms. This suggests that analyzing the effects of intentional peripheral visual field constriction on the sense of speed can aid in achieving driving control that is based on information associated with human visual perception.

Authors showed a design of scanning on the front and lateral sides of an automated guided vehicle (AGV) using a two-dimensional laser range scanner and mirrors. In this study, we used the scanning method shown by the authors, and we showed analysis of the scanning on the front and lateral sides of the AGV using the two-dimensional laser range scanner and mirrors. The mirrors are installed at the front of the scanner, side of the scanner, and back of the scanner. The scanner, mirrors and AGV are included in a rectangular area. The analysis shows to compare a rectangular area with the scanner, mirrors and AGV. The mirror has inside and outside edges away from the scanner. The width is from the outside edge to the outside edge in the two mirrors. The length is from the front of the scanner to the back of the AGV on the central axis of the AGV. The authors showed analysis method using the outside edge in the mirror, and the length and width of the rectangular area including the scanner, mirrors and AGV. Based on the analysis method, the authors showed results of the analysis using the sizes of an actual scanner, mirror and AGV. Used AGV was small, thin and large shape.

Unmanned aerial vehicles (UAVs) are expected to be used in various industrial fields. Therefore, for avoiding collisions with unmanned aerial vehicles against obstacles and other aircraft,the high-agility reference trajectory and control system of the Immelmann turn flight for a small unmanned helicopter which reverses the direction of travel by longitudinal turn are designed. The reference trajectory is derived based on mechanical and kinematic dynamics by dividing the a series of flights into four flight process and setting start and end conditions at each boundary. The design parameters including the start and end conditions and the time of each flight process can be given. For attitude control, design techniques based on the modified Rodrigues parameters are applied to design.And also,the variable gain dependent on the flight process is introduced to the velocity-control design. The flight tests using a small electric helicopter are condected, and it showed that the predetermined flight can be realized by using designed trajectory and control system in the real environment.

In order to develop safe and efficient overhead cranes, we developed an estimation technology for rope tension and load that can be applied to the initial sway suppression and overload detection during lifting from ground. To estimate the state of hoisting motor with nonlinear properties, we adopted multi-task learning using deep learning. The driving frequency, the rotation speed, and the current of the motor were input to the neural network, and the rope tension and the load were estimated simultaneously. The training data used for learning was created by image analysis of the movement of markers attached to load block and suspended load. The training and the validation data were prepared separately. To verify the effective of the learning results, we added data with different horizontal position and lifting length for the validation data. As a result, it was clarified that the tension condition of the rope and the load can be estimated simultaneously. The average delay of rope tension detection was 0.02 seconds, and the load estimation error was less than 7% of the maximum load.

In a previous study, the authors have clarified that the traffic volume of parallel residential areal roads increases during the time when the length of traffic jams becomes long on the main road during the morning and evening commuting hours. In this study, we investigated the section where traffic jams occur during the morning and evening commuting hours of National Highway No. 2 in Fukuyama City, Hiroshima Prefecture, and the residential areal roads parallel to this section. Then, the authors verified that the residential areal road is a loophole of National Highway No. 2 by performing a traffic route analysis using ETC2.0 data. In addition, by analyzing the traffic accidents that occur on this residential areal road, we clarified the dangers hidden in this road used as a loophole.

Many autonomous vehicles use LiDAR (Light Detection and Ranging) as external sensor to recognize surrounding environment. LiDAR sensor enables high resolution measurement by the emission of the laser scan lines. While in case of it is blocked by the obstacles, the sensor can’t measure the backside regions of it, and it is the measuring loss which is called occlusion. It makes the robot motion in unknown environment unsafe so pre-measured map has been used to let these robots recognize environment and to control the motion of it in previous studies. This study proposed an occlusion prediction method for speed control without the pre-measured map in unknown environment. This method is based on GAN (Generative Adversarial Networks) which is a type of deep learning model to predict the occluded regions from LiDAR measurement as local map for the robot. As a result, we showed applicability of this model for the robot speed control using the 3D robot simulator without the pre-measured map in unknown environment.

In order to provide driving support that matches the driver's intentions, it is desirable to predict the driver's intended driving behavior in advance. However, no matter what method is used, there is a problem of incorrect prediction. In this study, in order to avoid incorrect prediction of driving behavior, we investigated whether the characteristics of correct and incorrect prediction can be captured by visualizing the probability of driving behavior prediction with histograms. In this experiment, pressure distribution and driving operation information were input to a Bayesian Convolutional Neural Network (BCNN), and the parameters of the trained model were sampled multiple times to obtain multiple prediction probabilities, and a histogram of the prediction probability of driving behavior was obtained. As a result of the experiment, when the driving behavior was "Going straight", "Turn right", or "Turn left", there was a difference in the relative frequencies of the low and high prediction probability classes between correct and incorrect predictions, suggesting the possibility of correct and incorrect prediction classification based on the characteristics of the histogram of prediction probabilities.

In this study, we investigated and analyzed the correlation between high beam and low beam and irradiation characteristics using a headlight tester based on each inspection standard for automobile headlights. From the experimental results, it was confirmed that the irradiation characteristics of switching to the low beam do not fully satisfy the inspection criteria with respect to the appropriate irradiation direction and luminosity of the high beam. In particular, the clarity of the cut-off line is important for low beam measurement, and the elbow point cannot be measured accurately without a certain luminous intensity, and there is a concern that an inappropriate irradiation direction may lead to poor visibility. Therefore, the optimum adjustment position and low-luminosity irradiation characteristics were examined.

For improving safety and work efficiency in payload transporting using a hoist-type overhead crane, we investigated the control technology for suppression of payload-sway. In this report, we proposed a new control method to suppress the initial payload-sway caused by the position gap between the payload immediately above and the trolley that hoists and moves the payload, when the payload was eft from a ground. To suppress the initial payload-sway, the trolley was indeced above the payload using a rope length of the tension without slack and the posiiton gap was reduced. We proposed a asymptotic approach and a quadratic curve fitting. In the asymptotic approach, the trolley was moved toward a shorter rope length. In the quadratic curve fitting, the position immedeately above the payload was identified by fitting the trolley position and the rope length to the quadratic curve, and the trolley was moved to the identified postion. The proposed methods were compared by simulation, it was confirmed t that the performance was better in the quadratic curve fitting.

The feasibility of a low-temperature operable electric power system for space environment exposure experiments was discussed from the viewpoint of a secondary battery. After exploring various types of storage batteries, we evaluated two kinds of lithium-ion batteries (LiBs) as a candidate, and confirmed that one of them, LiB-B, can be charged with 0.1C rate at a low temperature of -20°C, and maintained more than half of its capacity even at -45°C with 0.2C discharging, besides, it can be operated stably at an extremely low temperature of -50°C. Such high performance of the LiB-B reveals the possibility to construct an electric power system for exposure experiments with minimal assistance, such as a heater or an insulation material. It is also expected to be applicable to power systems of microsatellites and other spacecraft. Moreover, we also suggested some materials for thermal control to improve the battery operation and its reliability according to the above-mentioned characteristics of LiB-B.

In this experiment, the aerospike engine combustion gas was visualized using the Schlieren method. Propellant were hydrogen and oxygen gas. Though the combustion gas choked at the throat, the shock wave was not observed around the nozzle.The characteristic exhaust velocity was evaluated, using the estimated propellant flow rates. C^*efficiency was evaluated to be 0.997, using the calculation result by NASA CEA.

The rocket engine combustion instability was investigated. The pressures increased by combustion were different in the previous two tests. Pressures were 1.5 to 3 times higher than that before combustion in 2019 model. In another 2020 model, pressures were 1.1 to 17 times higher. In the present study, the equivalence ratio in the two models was measured. Fuel was 2-propanol, and oxidizer was oxygen gas. Mixture was sampled from the combustor models. According to the sampled gas, the equivalence ratio was approximately 0.7 in 2019 model. In 2020 model, the ratio was approximately 1. According to mixture temperature in the combustor model, mixture was vapor in 2019 model, and was mist in 2020 model. The large equivalence ratio and large pressure increase will have relation to the mixture condition.

In the shock testing of spacecraft components, alternative methods are used where the shock source is different from the actual one, such as the mechanical impact method. Adjusting the shock response spectrum (SRS) in these tests to desired values is a difficult task that requires considerable expertise. Therefore, we propose a new SRS adjustment concept in which a structure is inserted between the base plate of the shock tester and the test specimen. This method alters the propagation characteristics of the vibration into the specimen. Due to its easy implementation, this method can also be applied to other mechanical shock testers in use. And this method has good affinity with the substructure synthesis method and can predict the test level in advance. This paper presents the prediction of shock test levels using transfer function synthesis method.