Human perceptual and affective responses change dynamically when stimuli are experienced. In this study, we developed a method for modeling the causal structures of affective dynamics using time-series data. Using the temporal dominance of sensations method, perceptual and affective data were collected from individuals eating strawberries, and the resulting time-series data were mathematically represented using a vector auto-regression model. Multihierarchical and multidimensional causality structures that explain the temporal evolution of perceptual and affective responses were then established based on Granger causality and the information criterion. The established model suggests how affective and preferential responses are triggered following exposure to stimuli. We also assessed the quantitative and semantic validity of the model.

Temporal dominance (TD) methods can be used to record temporal changes in multiple sensory and affective responses. TD methods have found wide applications in the analysis of eating experiences of humans. However, extant analyses performed on TD data do not fully utilize the time-series properties of such data. The present study validates the prospect of principal motion analysis (PMA) of TD data. PMA is an extension of principal component analysis, and can be used to resolve multivariate motion data into base principal motions. In this study, panelists were asked to evaluate the tastes of ten types of pickled plums using the temporal dominance of sensations (TDS) and emotions (TDE) methods. Additionally, the panelists were asked to rate the plums using the semantic differential method. Results obtained using both methods were observed to demonstrate good agreement with each other in terms of the structures of reduced variable spaces. As realized in this article, implementation of the combined TD–PMA approach can potentially facilitate statistical discrimination of all food products, whereas conventional methods, such as principal component analysis of data provided via use of the semantic differential method, can at best discriminate only 67 percent of product pairs. PMA can, therefore, be considered as a suitable technique to reveal the characteristics of TD data.

There is increasing demand for tactile feedback functions for touch panels. We investigated whether virtual roughness texture quality can be improved through simultaneous use of vibrotactile and electrostatic-friction stimuli. This conjunctive use is expected to improve the perceptual quality of texture stimuli, because vibrotactile and electrostatic-friction stimuli have complementary characteristics. Our previous studies confirmed that these conjunct stimuli yield enhanced realism for simple grating roughness. In this study, we conducted experiments using simple and complex sinusoidal surface profiles consisting of one or two spatial wave components. Three different evaluation criteria were employed. The first criterion concerned the subjective realism, i.e., similarity with actual roughness textures, of virtual roughness textures. Participants compared the following three stimulus conditions: vibrotactile stimuli only, electrostatic-friction stimuli only, and their conjunct stimuli. The conjunct stimuli yielded the greatest realism. The second criterion concerned roughness texture identification under each of the three stimulus conditions for five different roughness textures. The highest identification accuracy rate was achieved under the conjunct stimulus condition; however, the performance difference was marginal. The third criterion concerned the discrimination threshold of the grating-scale spatial wavelength. There were no marked differences among the results for the three conditions. The findings of this study will improve virtual texture quality for touch-panel-type surface tactile displays.

As touchscreens have become a standard feature in mobile devices, technologies for presenting tactile texture feedback on the panel have been attracting attention. We tested a new method for presenting natural materials using an electrostatic tactile texture display. In this method, the frictional forces are decomposed into low- and high-frequency components. The low-frequency component was modeled based on Coulomb’s friction law, such that the friction force was reactive to the finger’s normal force. The high-frequency component was modeled using an auto-regressive model to retain its features of frequency spectra. Four natural material types, representing leather, cork, denim, and drawing paper, were presented to six assessors using this method. In a condition where only the low-frequency friction force components were rendered, the materials were correctly recognized at 70%. In contrast, when the high-frequency components were superposed, this rate increased to 80%, although the difference was not statistically significant. Our approach to combine a physical friction model and frequency spectrum for low- and high-frequency components, respectively, allows people to recognize virtual natural materials rendered on touch panels.

In this study, improved realism of virtual roughness textures is verified by using vibrotactile and electrostatic friction stimuli in conjunction. Because these two types of stimuli have complementary features, they are expected to enhance the perceived quality of stimuli. In our previous research, conjunct stimuli improved the realism of virtual surfaces with sinusoidal surface roughness profiles. In this research, we tested with complicated surface roughness, and experimental participants ranked vibrotactile, electrostatic friction, and conjunct conditions in terms of realism compared to actual roughness textures. The conjunct condition was judged to be more realistic for simple and complicated surface profiles consisting of one or two spatial wave components.

In recent years, the number of patients suffering from diabetes mellitus has continued to increase. When diabetes becomes severe, ulcers may form on the feet of the patient. In the past few years, several researchers have focused on the risk factors and avoidance of ulceration. One effective method to predict the occurrence of diabetic foot ulcers is developing a machine learning model. However, few studies combine both clinical indices and mechanical data as the attributes of the training datasets. In this study, we developed a composite model of a convolutional neural network and K-means clustering to extract features from diabetic patients with or without ulceration as well as healthy individuals. Using a deep clustering model, the center of pressure (CoP) trajectory images were divided into three clusters. Furthermore, we evaluated the performance of the clustering by extracting the features from the CoP trajectory images in each cluster and combining them with the clinical indices of the patients. The results showed that patients with ulcers when walking tend to contact the ground with a narrow area of the plantar and apply a small force. Furthermore, it was found that patients undergoing diabetic neuropathy or with a toe amputation have a high potential of suffering from ulcers.

Many manufacturing industries have a high demand for the construction of collaborative operation systems using industrial robots. Although there is a preexisting set of safety verification data in ISO/TS 15066 for collaborative operations, there is no established testing method for safety validation. To establish a testing method, it is effective to use a dummy that has mechanical properties similar to those of a human. However, there is no parametric study that exists for designing a dummy that represents the static and dynamic mechanical properties and the nonlinearity of the static mechanical properties. In this study, static and dynamic experiments were conducted to obtain the mechanical stiffness of human subjects and the contact force transitions during the dynamic contact between a robot system and a human. Subsequently, the same experiment was conducted using the proposed dummy. The biofidelity of the dummy was examined by comparing the parameters of a viscoelastic model. This study contributes to increasing the safety of collaborative operations by developing a dummy that can be used for risk assessments in collaborative operations.

Although a wearable assist robot helps to enhance the gait ability of the wearer, it can cause gait instability in an emergency. A collision with an environmental object is one source of such instability. Insufficient adaptation to change by the physical frame, and restriction of joint motion, increases the risk of collision and fall of the wearer owing to the robot. In this article, the reaction motion owing to side contact, which applies a spin moment to the body, is investigated for the wearers of a physical assistant robot. In particular, reaction motions in the horizontal plane, such as body rotation, foot direction, and center-of-mass position, are investigated. Factor analysis and cluster analysis are performed, and two reaction patterns-the rotation group and the straight group-are determined. The motion of these groups suggests an approach for navigation through an obstacle while using an assist robot. Furthermore, the gait phase at collision time is considered as a determinant of the reaction pattern and the reaction pattern is found to affect the fall mode. The results of the article suggest that the joints of the physical assistant robot should be equipped with sufficient degrees of freedom, at least in hip rotation, such that it is capable of rotating the body and feet in the horizontal plane to reduce fall risk.

Vibrotactile and friction texture displays are good options for artificially presenting the roughness and frictional properties of textures, respectively. These two types of displays are compatible with touch panels and exhibit complementary characteristics. We combine vibrotactile and electrostatic friction texture displays to improve the quality of virtual textures, considering that actual textured surfaces are composed of both properties. We investigate their composition ratios when displaying roughness textures. Grating roughness scales with one of the six surface wavelengths are generated under 11 display conditions, and in 9 of which, vibrotactile and friction stimuli are combined with different composition ratios. A forced-choice experiment regarding subjective realism indicates that a vibrotactile stimulus with a slight variable-friction stimulus is effective for presenting quality textures for surface wavelengths greater than or equal to 1.0mm.

Estimating the physical burdens of people with motor dysfunction is helpful for developing care equipment and facilities. We propose a patient simulation where a trained healthy subject simulates impaired movement with instruments to guide his/her body motion with less possible incurred pain. Electric cutaneous stimuli are used to intuitively indicate possible pain of an injured body part. We used this simulator to investigate a suitable handrail position for knee pain holders during sit-down and stand-up motions. Electric stimuli were presented in proportion to the moment at the supposedly impaired knee. The results suggest that a handrail on either the healthy or impaired side significantly reduces the experienced physical burden (i.e., internal knee flexion-extension moment). Specifically, the maximum knee moment was smaller when the handrail was on the impaired side than it was on the healthy side.

Humans recognize textures through the interaction induced by the normal and shear forces generated between the fingertip and the surface of objects when sliding their fingers on material surfaces. We investigated the stochastic relationships of these two axial forces during tactile exploration. The characteristic values of the elliptical distribution obtained by the variations of the two axial forces were analyzed. It was found that the sizes, shapes, and gradients of the elliptical distribution of the low and high frequency forces statistically varied among the 18 types of textures. These parameters may be the indicators to distinguish the unevenness of the surface profile and the behavior of friction between the surface of material and the fingertip. The fact that the stochastic relationships between the two axial forces depend on the surface of material help us understand human texture perception.

Forward fall is one of the most common causes of upper extremity fractures. Significant factors influencing impact force and injuries were widely studied; however, it is necessary to investigate the natural reactions of humans during a forward fall to obtain a realistic evaluation of injuries. The purpose of this study was to analyze the natural motion of the upper extremity during an induced trip. We carried out a tripping experiment using an obstacle colliding with one leg; while recovery step was prevented to produce a forward fall. Results showed that the elbow extension had a slight ascending trend during the forward fall and elbow angle at the moment of hand-ground contact was appropriate to reduce the peak force. Landing on the obstacle-side hand was more likely due to body rotation towards the obstacle-side. To prevent injuries, subjects were connected to a safety harness not to strike the ground with high impact velocity. Thus, the fall motion was simulated using a 12 DOF model to obtain a realistic evaluation of the impact velocity and the related impact force caused by the forward fall was estimated using a sagittal 3-segment model. Results of this study can be useful in human-robot collaboration, where a collision between human and robot may cause a forward fall.

Many wearable robots are not fully capable of fitting the motion of the wearer, owing to their limited degrees-of-freedom (DOF). This limitation disturbs motions such as corner curving, which are required in daily life. In this study, the effect of the DOF restriction on the corner curving motion was observed and analyzed. The gait motions, when curving a round corner with and without the restriction of out-of-sagittal plane motion, were compared using a wearable device which could restrict hip adduction/abduction and rotation. The result suggested that the restriction not only decreased the range of motion of hip adduction/abduction and rotation, but also changed the center of mass (COM) trajectory. Owing to the disturbance applied during the stepping motion of the outer leg, the COM trajectory moved further from the corner in the trials with restriction. Thus, the distance between COM and the base of support of the inner foot became larger during the swing phase of the outer leg, which possibly increased the risk of loss of balance. Furthermore, it was also suggested that an assistive torque, which does not consider curving motion, possibly increases such risk.

This study demonstrates that the simultaneous use of two types of representative principles for tactile texture displays, vibrotactile and electrostatic friction stimulation surpasses either type of stimulation. We used a set of sinusoidal grating scales with different surface spatial periods as simulated textured surfaces for testing the quality of texture displays. Experimental participants judged the best stimulus condition among the vibrotactile, electrostatic friction, and the conjunctive stimuli in terms of the reality perceived from the simulated texture stimuli. The use of vibrotactile and electrostatic stimuli in conjunction was judged to be the most realistic for the spatial periods of 1 and 2 mm, whereas noticeable differences among the three stimuli conditions were not reported for small spatial periods of 0.2 and 0.5 mm. The simultaneous use is effective for simulating the surfaces with middle-level asperity.

Equinovarus is a foot deformity characterized by the patient's foot being at rest in an abnormally supinated state. In a clinical setting, physical therapists manually stretch the foot to a pronated state to allow the patient to retain some degree of mobility. Currently, no automated ankle stretching machine is commercially available. To tackle this issue, we have developed an automated stretching machine that controls the eversion and flexion angles of a patient's deformed foot using two pneumatic actuators. We designed a proportional and integral (PI) controller to place the foot in the desired dorsiflexed position and performed a user test involving a healthy participant. Even when the initial foot position of the participant was in an equinovarus position, the foot was successfully everted and dorsiflexed to match the desired reference posture via the stretching machine. The average differences between the reference and measured foot angles at the final state were found to be within 2° for both the dorsiflexion and eversion angles. The machine replicated the reference angles with acceptable errors.

The elderly sometimes fall owing to tripping caused by decreased dorsal flexion during the swing phase. A wearable assist device that was able to increase dorsal flexion to reduce the risk of tripping was proposed in this study. This device assists in dorsal flexion during the swing phase by transferring the tension force from the upper body to the foot by connecting them using strings. To improve the efficiency of the assist, the string path was optimized to maximize the assist torque at the time of minimum foot clearance (MFC). In this study, two string-path patterns were tested. Based on the walking experiment results, we confirmed that the wearable assist device that we developed successfully increased both the dorsal flexion and minimum foot clearance during the swing phase.

Three-dimensional stretching is needed to treat equinovarus, which deforms the patient's foot to plantarflexion, adduction, and inversion postures. We have prototyped a three-dimensional stretching machine that the patient can use for the treatment of equinovarus by him- or herself. By adopting a cable-driven mechanism with two independently controllable pneumatic actuators, the stretching machine can apply a force to the foot along the dorsiflexion direction as well as the direction combining abduction and eversion. In this study, we verified the effectiveness of a prototype stretching machine for healthy subjects. For evaluation, the muscle stiffness and maximum voluntary contraction (MVC) of plantarflexion were compared immediately before and after stretching and 10 min later. As a result, the MVC decreased after stretching, which is a clinical index for effective stretching.

The estimation of the physical burdens from which people with motor impairment suffer helps us establish welfare techniques comprising personal care equipment and assessment of critical risks, such as fall risks. However, the involvement of actual patients in the evaluation and development of this equipment is costly and involves the exposure of patients to long and exhausting experiments. To solve this problem, we developed a robot wearable by a healthy person and the associated control algorithm to simulate typical motions of patients with knee osteoarthritis, which is a common symptom for the elderly and causes pain during movement. To estimate the physical burdens inflicted by knee malfunctions, we computed the knee flexion and extension moment of the simulated patient during the standing-up and sitting-down motions. The moments, estimated under certain conditions, are qualitatively consistent with those considered clinically, which corroborates the validity of our patient simulation techniques.

We developed a method to present an impulsive resistance force by using the passive damping brake of a DC motor. The rapid pulsive change in the resistance forces preceding the main brake increased the perceived resistance of the brake. In a ranking task involving five naive participants, a damping brake with two preceding pulsive resistances was reported to deliver a resistance larger than that due to a simple stepwise brake that achieves the maximum physical braking force. This finding indicates that perceptually effective impulsive resistances are different from the physically optimal resistance. Our technique enhances the abilities of passive force displays, which are inherently safe human-machine interfaces.

It is ethically problematic to engage people with impaired motor functions in experiments, wherein they may face severe pain and unacceptable risk of injuries. We developed a wearable robot for healthy adults and its control algorithm to simulate the behaviors of patients with knee osteoarthritis. We focused on invariant motions, which are typically presented for avoiding knee pain. In order to simulate patient movements by a healthy person wearing a robot, we formulated a model motion that represented invariant patient motions. We then determined the output torque of the exoskeletal knee robot such that the wearer's motion followed the model motion. The effectiveness of the method was testified for standing-up and sitting-down motions, and some characteristic impaired motions such as body inclination to the healthy side and imbalanced right and left knee angles were manifested by the simulator.

Physical assistant robots are expected to be able to assist in our working and daily lives. However, safety problems arise when robots are used in indoor environments with obstacles. The focus of this study was on collision of the side of a robot with an obstacle, and the recovery motion after the collision was measured. A healthy young man participated in this experiment. Two different types of recovery motions were observed. These two motions differed in the degree of deceleration of the body rotation in the horizontal plane. The part of the robot and the timing of the collision with the obstacle affected the recovery motion strategy. A collision at the hip side in the middle swing phase might be the most severe condition from which to recover, among those considered in this study. The results of this study will be helpful in improving the safety of the use of physical assistant robots.

Humans can perceive the hardness of an object from the damped natural vibration produced by tapping its surface. This damped natural vibration can be used in hap tic rendering to effectively present a sense of the hardness of an object. Although its frequency is known to influence the subjective hardness, the effects of the viscosity or decay rate have yet to be studied. We researched the contributions of the mass, viscosity, and stiffness to the hardness perception using a commercial force display and psychophysical experiments, where only the vibratory displacement was presented. As a result, the stiffness was found to have a positive effect on the subjective hardness, similar to the usual hap tic rendering of an object. The mass of the object was negatively correlated with the subjective hardness. These results indicated that the vibratory frequency was the dominant parameter in determining the perceived hardness, which agreed with the previous reports on the perceptual effect of damped natural vibration. In contrast, the perceptual effect of the viscosity depended on the individual. For approximately half the participants, the viscosity did not directly influence the subjective hardness of an object. Some participants felt that a damped natural vibration with greater viscosity was harder. However, the other participants felt that it was harder with a smaller viscosity, which suggested that the decay rate of the mechanical system was used as a criterion to judge the hardness of an object. Therefore, the results indicated that the vibratory frequency is a general parameter that can be used to specify the subjective hardness, whereas the perceptual influence of the viscosity or decay rate depends on the individual.

Falls are one of the main causes of decreased quality of life and activities of daily living among the elderly individuals. Because tripping is the most common cause of falls, some types of recovery motions have been reported after tripping induced by a flat plate or block. Common recovery strategies observed under these conditions included elevating and lowering strategies. Especially, in case that the swinging leg contacts an obstacle at its early phase, the elevating strategy has been considered as a typical recovery motion. In this study, we performed a tripping experiment using a crossbar in the early and middle of swing phase. As a result, a new recovery strategy, which differed from those previously reported, was observed. In the new strategy, referred to as the skip strategy, the subjects lifted and moved the standing leg forward after tripping. The subjects utilized this strategy to recover from an early phase trip. The participants had to exert enough knee flexion and ankle plantar flexion torques of recovery leg with the skip strategy and enough knee extension, hip extension, and ankle plantar flexion torques of recovery leg with the lowering strategy. Especially for the elderly who cannot perform large joint torques, gait assistant devices such as physical assistant robots are required to compensate such joint torques in order to perform these recovery motions.

Measuring the shear deformation of a finger pad during active hap tic exploration of materials is important for analysis of textural sensations and development of tactile texture displays. Thus far, there has been no general methods to measure such deformations for active touch. To create a sensor system, we have been developing a new method for estimating the deformation of a finger pad based on the skin deformation propagated to the radial side of a finger tip. In this study, in order to validate the method, we compared the deformation or its acceleration of finger pad estimated on the basis of the acceleration measured at the radial side and those estimated by the shear force applied to the finger pad while exploring a few types of materials. The estimation errors for the deformations at 40 -- 450 Hz were smaller than human discrimination thresholds, indicating that the accuracy of our method is satisfactory compared with human perceptual sensitivity.

We have developed a texture display system that modifies the perceived roughness of textured surfaces via a voice coil actuator worn on the finger. To increase the roughness sensations, the vibrotactile stimuli from the actuator simulate the skin deformations that are activated when a wavy surface is scanned. Conversely, to decrease the roughness sensations of the textured surface, a high-frequency vibrotactile stimulus offsets the activity levels of tactile mechanoreceptors. This offset suppresses the perceived roughness of the surfaces being touched, with increase in the offset correlating with increase in the feeling of smoothness of the surfaces. We conducted an experiment in which we tested the effects of these two types of vibrotactile stimulation on grating roughness specimens, with subjective responses acquired from eight participants via the magnitude estimation method. The results obtained indicate that our method selectively increases and decreases the roughness felt. The intention is for the technique to be used to develop an augmented reality device for textures.

Risk assessment considering child injury risks in collisions should be performed before introducing mobile robots into human-robot collaborative environments. This paper provides data for estimating the injury level and the injury probability for risk assessment using information from automotive-accident research. Experiments representing collisions between six-year-old children and mobile robots were conducted using an automotive crash-test dummy. Mechanical data on the head, neck, and chest of the dummy were measured. The results indicate that robot manufacturers should consider head injuries in collisions before introducing their mobile robots into environments with children.

Augmented reality technology realizes variable sensory feedback while maintaining a high perceptual quality by superposing real and virtual stimuli. In this research, we experimentally verified whether the masking effects of high-frequency vibrotactile stimulation on fingers is applicable to augmented reality of the tactile texture of materials. A vibrotactile stimulator was worn on a finger (ring-vibrating condition) or mounted on materials (material-vibrating condition). We investigated the effect of vibratory stimuli on perception of sand-paper roughness under these two conditions. Under both conditions, participants' performances to discriminate the roughness diminished because of the masking effects of the high-frequency vibratory stimuli (experiment 1). Subsequently, we specified how the masking effects influenced the perceived roughness using Scheffe's paired comparison. For the ring-vibrating condition, as accelerations of the vibrotactile stimuli increased, all participants reported a decrease in the perceived roughness of the sand-paper. For the material-vibrating condition, the roughness percepts increased or decreased depending on the individual (experiment 2). Although both implementation methods affect roughness perception, the ring-vibrating condition is recommended for augmented reality of tactile textures because it can steadily control the decrease in perceived roughness of touched materials.

In this study, psychological experiments are conducted to investigate harm-avoidance action characteristics in humans in close contact with robotic devices. For the experiments, a situation is created in which the sharp end-effector tip of a robot suddenly approaches the eyes of a facing participant. The avoidance reaction time is defined as the time interval from the beginning of the end-effector motion to when the participant begins movement to avoid harm. The results suggest that the avoidance reaction time does not depend on the type of work being performed but on the initial distance between the human's eyes and the approaching object. Using this information, safe human-robot working distances can be determined.

For robots working in human environment, human safety is one of the most serious and important aspects to be considered. Appropriate intrinsic and / or functional safety design must be employed for collaborative robots in order to reduce the human injury risks through physical human-robot interaction. Above all the situations that such interactions are possible to occur, it is reported that clamping humans with robot and its environment is of great danger in a sense of human injury risks. We have conducted experiments and simulations that mimic clamping-human accident in some different situations varying robot's inertia, velocity, thickness of robot's soft cover and joint stiffness. Furthermore, some collision detection and reaction measures are implemented on the experimental system and their contributions to human safety were evaluated. From the results of these experiments and simulations, the joint stiffness was found to play an important role on human safety considerably affecting the performance of collision reaction.

Wearable robots have finally reached a point at which they can be put to more practical use. However, methods for assessing their safety have yet to be properly established. Furthermore, for ethical reasons, such assessments must be conducted without exposing the human user to unnecessary danger. Therefore, in this study, a lower-leg dummy for the safety assessment of a wearable robot was developed to evaluate the effects of a misaligned knee joint.

In this study, we developed vibrotactile display methods that can assist designers in product design. In order to achieve realistic sensations required for such designing purposes, we used real materials such as cloth, paper, wood, and leather and applied vibrotactile stimuli to modify the roughness sensations of these materials. This approach allowed us to present textures of various virtual materials with a strong sense of reality. We verified that our proposed methods could selectively modify the fine and macro-roughness sensations of real materials. The methods are expected to aid product designers in deciding tactile sensations suitable for their products.

In this study, a psychological experiment is conducted to investigate harm-avoidance action characteristics in humans in close contact with robotic devices. For the experiment, a situation is created in which the sharp end-effector tip of a robot suddenly approaches the eyes of a facing participant. We define three parameters that represent harm-avoidance action characteristics: the avoidance reaction time, maximum avoidance acceleration, and maximum avoidance speed. The results suggest that the avoidance reaction time depends on the initial distance between the human's eyes and the approaching object. The results show that there are individual differences in all parameters studied. The results also show that the avoidance reaction time is negatively correlated with the other two parameters, which have a strong positive correlation with each other. We conclude that the avoidance reaction time and maximum avoidance speed are considerable parameters on avoidance actions.

A vibrotactile texture display produces virtual textures by applying vibratory stimuli to finger pads. In this study, we developed a technique to alter such textures based on certain specified materials. For example, the technique allows us to alter vibrotactile textures using terms such as “wood-,” “cotton-,” or “paper-like” which are familiar to end users of displays. The altered textures feel more similar to these specified materials. We realized this technique by constructing a material space, in which the materials are located based on the features of their vibrotactile spectra. The vibrotactile textures were then modified in this space. Our experimental results show that the technique can be used to alter a virtual wood texture to feel like cloth.

In this paper, we studied the conditions require to ensure the safety and efficient traffic of pedestrians walking in an indoor blind corner-based decision model by determining their behavior. This model was constructed with consideration for tactics that affect when person encounter each other using a mixed strategy from game theory, which we expanded using the Dempster-Shafer theory of evidence for the adaptation of behavioral changes caused by information uncertainty. We determined the conditions required for safe and efficient traffic, and we verified their effectiveness in a simulation using a decision model. Based on the knowledge obtained from this verification, we tested an implementation of a system that supported safety and efficient traffic in an indoor blind corner-based model.

For researchers of haptic interfaces, evaluation of the perceptual similarity between virtual and real haptic stimuli has long been a serious problem. One of the most commonly employed evaluation methods is an identification task where assessors identify the type of randomly presented stimuli among multiple candidates. The results of this method are summarized as confusion matrices. We developed a method that allocates all virtual and real stimuli in a perceptual space. The spatial distribution of the stimuli allows us to visually understand the perceptual relationships between the stimuli. A brief validation confirmed that the proposed method is effective in evaluating the perceptual similarity between virtual and real stimuli.

In the design of human-robot coexistence environments, human avoidance actions are not taken into consideration. We conduct a psychological experiment to observe human avoidance actions where a mechanical harm approached his or her eyes. Using the observed motions, we carry out “extrapolation simulation” of human avoidance actions in collision situations. As simulation results, first, we express collision probability using the motion range of the end effector of a robot as a variable. Second, we identify the motion conditions, which ensure the safety, of the end effector in our setup. Finally, we quantitatively compare the safety conditions between taking and not taking avoidance actions into consideration. This study contributes to providing a tool for conducting a practical risk assessment aimed at realizing a safe human-robot coexistence system.

For reasons that have not yet been investigated, some textures appeal to human touch. We investigate the relationships between the physical and sensory factors, the predictability of textures, and their appeal to human touch. We conduct sensory evaluation of 24 artificial clay textures, and the results are analyzed using factor analysis. Experimental results identify five sensory factors. We reveal that 70-80% of the textures' appeal to human touch can be described on the basis of the sensory and physical factors. We propose two models involving the predictability of textures. These models, as well as models based on the sensory and physical factors, explain the affinity to textures.

The presentation of textures is a major application of vibrotactile displays. Lossy data compression is an effective approach for reducing a data size of such textures when delivering them through the Internet. This study developed an objective index that well described material-like textures' quality changes caused by data compression processes. We introduced a linear combination of the power and amplitude spectra that were weighted by the inverse of a detection threshold curve of vibrotactile stimuli. This combination described the quality deterioration of compressed textures to an accuracy of approximately 80%. These objective indices enhance the development of the lossy data compression algorithms without psychological evaluation tests of subjective quality changes in the textures.

The objective of our study is to develop a safety measure that ensures smooth control mode switching of an intelligent assist device (IAD)-Skill-Assist for the purpose of effective automotive manufacturing. Smooth switching, i.e., switching without abrupt deceleration, from the hands-off control mode (used for automatic travel) to the hands-on control mode (as typified by a power-assist control mode) is expected to enhance the productivity of IADs. However, the hazardous behaviors, i.e., careless behaviors that disregard an approaching IAD in hands-off control mode might cause collision and serious injuries to a human operator. For safety purposes, we focused on the operator's gesture of reaching for the force sensor of the Skill-Assist. Laser range sensors were employed to recognize the reaching gesture during control mode switching. A reaching gesture recognition algorithm, based on a hidden Markov model, processes characteristic patterns acquired by the laser range sensors, and judges the operator safety. Experimental results suggest that the Skill-Assist to which the proposed safety measure was applied could recognize the reaching gesture of an operator when the control mode was smoothly switched. In an experiment on hazardous behavior, the Skill-Assist could predict that the operator would be unsafe.

This paper presents a new design method for force control, which aims to control the interaction force between a robot and a human by satisfying certain safety requirements. In this method, an optimization-based control algorithm, called Iterative Feedback Tuning (IFT), is used to employ safety requirements as constraints of an optimization problem, which is then solved using sequential quadratic programming (SQP). Therefore, this control method is applicable for safety-critical systems such as personal service robots. These robots are developed to provide assistance to patients or disabled people in their daily life by performing human-robot contact tasks such as wiping the face with a towel, scratching, etc. In addition, in IFT, a Newton search direction to update the controller parameters at each iteration is obtained on the basis of the closed-loop experimental data. Hence, IFT does not require explicit modeling of environment, in particular, human dynamics, otherwise in model-based approach it will be hard task to obtain a useful model. In the simulation and experiment, the effectiveness of the proposed method is examined by applying it to 1-DOF contact system.

A direct teaching system emphasizing safety and operability for SP02, an upper-body humanoid that cooperates with a human in a cell-production workspace has been described. The system has a direct teaching device, which enables an operator to teach a multi-D.O.F manipulator at comfortable posture despite large changes in the pose of an end-effector. The double-checked safety-related part(SRP/CS) introducing a safety module satisfies the safety performance defined in the international safety standard. To make the teaching easier to perform, impedance parameters in a variable impedance control are set through several direct teaching experiments. Through a simple direct teaching experiment, the safety performance of the SRP/CS and the teaching performance of the direct teaching system was successfully confirmed.

Although the need for humans and robots to work together in close proximity is increasing, this is currently not allowed with conventional industrial robots for reasons of safety. Next generation robots that can work safely in close proximity with humans must not only be highly functional, but must also be highly reliable with built in safety features. With this aim in mind, we have developed a safety module that integrates safety functions required for robots to work side by side with humans. The safety module is designed to be compliant with international safety standards and Japanese law. Redundant sensory signal processing by an external hardware module and plug-in software installed in the robot controller ensure high-reliability and flexibility. This paper describes the concept and design of the safety module and shows some evaluation results of its safety functions.

It is believed that from this point forward, there will be a need for industrial robots that work alongside and cooperatively with humans. However, the current development of existing robotics technology is inadequate to ensure the safety of such new industrial robots. Our research proposes a safety-planning technology called Coexistence Hazard Avoidance Technology, designed for use with a low-powered human-collaborative industrial robot. We load this technology into a risk-management simulator and verify that by using it, a dynamic planning method for the safe operation of robots can be reasonably undertaken in terms of both theory and calculation.

The objectives in this paper are to develop a method for safety-preservation oriented reaching monitoring (SORM) and to demonstrate its effectiveness for realizing a safe and smooth control mode switching. For more effective manufacturing activities, smooth control mode switching of a power assist device ldquoSkill-Assistrdquo from hands-off to hands-on is expected to enhance the productive efficiency. However, hazardous behaviors of an operator without paying any attention to a Skill-Assist in the hands-off control mode are possible to occur, and may result in serious injuries; that is why the SORM is demanded for smooth control mode switching. We focused on operator's hand reaching to the force sensor of a Skill-Assist conducted in the control mode switching, and two laser range sensors are employed to monitor the reaching. The SORM algorithm based on hidden Markov model processes the reaching information, and judges predictively whether the operator will be in a safe state or not. An experimental result shows the Skill-Assist to which the SORM is applied could recognize a control mode switching behavior with reaching while the control mode is switched smoothly without rapid deceleration. In an experiment for a hazardous behavior without reaching, the Skill-Assist could judge successfully an operator will be in a unsafe state. The judgment in the experiment for the hazardous behavior was evaluated by a minimum safe distance specified by an ISO safety standard, and it could satisfy the safety requirement.

To develop a nonverbal communication channel between an operator and a system, we built a tracking system called the Adaptive Visual Attentive Tracker (AVAT) to track and zoom in to the operator's behavioral sequence which represents his/her intention. In our system, hidden Markov models (HMMs) first roughly model the gesture pattern. Then, the state transition probabilities in HMMs are used to assign as the rewards in temporal difference (TD) learning. Later, the TD learning method is utilized to adjust the action model of the tracker for its situated behaviors in the tracking task. Identification of the hand sign gesture context through wavelet analysis autonomously provides a reward value for optimizing AVAT's action patterns. Experimental results of tracking the operator's hand sign action sequences during her natural walking motion with higher accuracy are shown which demonstrate the effectiveness of the proposed HMM-based TD learning algorithm of AVAT. During TD learning experiments, the exploring randomly chosen actions sometimes exceed the predefined state area, and thus involuntarily enlarge the domain of states. We describe a method utilizing HMMs with continuous observation distribution to detect whether the state would be split to make a new state. The generation of new states brings the ability of enlarging the predefined area of states.

The goal of our study is development of a "tactile mirror" for displaying tactile perception information. A point of distinctive difference between a tactile mirror and conventional tactile displays is that a tactile mirror possesses a feedback mechanism for acquiring the contact state information on the finger pad concerning not force but tactile perception sensation. The primary development concerns human static friction sensation. First, a control method and the structure of a tactile mirror are described. Second, simulation results using FEM for contact analysis and a ANN(artificial neural network) for generating feedback signals are shown. Finally, experiments demonstrating that providing a feedback component accomplish accurate display of static friction sensation with a conventional type of tactile displays.

It is considered that kinesthetic feeling induced by the kinetic characteristic of a manipulated object plays an important role for performance of human daily skills. Therefore, if we consider a wearable robot/joint as the robot/joint operated by the daily skills, it is necessary to control the robot/joint so that kinesthetic feeling induced by the kinetic characteristic of an object is directly transmitted to a wearer, without any distortion by the kinetic characteristic of the wearable robot/joint itself. This is the control issue of "kinesthetic transparency". This paper discusses a method of control to improve kinesthetic transparency of a wearable joint. Kinesthetic transparency can be basically improved by active compensation. In order to avoid necessity of offline estimation for unknown kinetic parameters of the joint, this paper formulates a control law which automatically estimates the unknown parameters during operation of the joint. Simulation results demonstrate effectiveness of the method. However, experimental results reveal that accurate compensation of frictional terms is necessary for further improvement.

When we use a tactile sensor, sensing mechanism and restoration of texture from electric sensing signals are important issues. The objectives of this research are to design a new texture sensing system and to develop a new signal processing algorithm which can restore various texture. The new texture sensing system is designed to get texture with high resolution and wide velocity range, which uses a PVDF sensor and has fixing components for several types of objects in precise condition. Next, a new signal processing algorithm is developed to restore texture. In the previous researches, forward model is needed and then it is inverted by using some regularized inversion formula in frequency domain, where there exist problems such as amplification of noise due to ill-posedness, modeling uncertainty due to orientation and position of PVDF films and silicon rubber, and difficulty to add nonlinear terms into model in frequency domain. While, in this paper, we model directly the relation containing transient-state from measured signals to texture by using model structure of multi-input multi-output nonlinear autoregressive moving average and a time domain least squares estimation. The direct modeling can be done by the use of F/T sensor which is not used in the previous researches. Finally the several texture is experimentally reconstructed from sensing signals using the developed signal processing algorithm.

To successfully incorporate human skills into a wearable robot, it is required to design its joint mechanism so that it does not interfere with the human skills. From this viewpoint, this paper first addresses six design goals for the joint mechanism. Based on the result of preliminary examination for two well-known mechanisms, a new 6-DOF mechanism referred to as the Wearable HEXA mechanism is proposed. This is based on the HEXA, which has been originally developed as a high-speed manipulator. Detail of its design process is given, and evaluation of a trial mechanism manufactured along the design process is also reported.

We propose a method for preventing hazardous accidents due to operator's action slip in their use of skill-assist, which has been already introduced as a power assist device in automobile assembly processes. First, we show that hidden Markov model (HMM) can be used to detect human erroneous operation from data sequences of an operator's hand motion trajectory, but that a problem arises in direct application of HMM to human error detection when trajectory is not in alignment with any of pretaught trajectories. However, HMM-based Dempster-Shafer theory proposed in the study allows the system to select their safety- or productivity- oriented robot control policy, solves the detection problem, and performs secure accident prevention, Second, a workability improvement process which comprises a state-policy and a teach data renewal sub-processes allows for optimal reconstruction of the policy determinant state space and HMMs. Finally, experimental results verify our overall proposal, and demonstrate the effectiveness of the workability improvement process.

In the social context of an aging workforce in automobile final assembly processes, modular assembly systems have become a dominant trend to reduce the workers' physical workload. Although a modular assembly system provides various benefits, modularized components are heavier and larger than their preassembled parts. As a result, tasks of installing modular components have become increasingly more difficult and laborious for the workers. Though automated assembly processes have been introduced in some cases, systems with such processes are ineffective because they are not flexible enough to allow manufacturer to respond to rapidly changing demands. In an effort to solve the problem, we have developed a new type of human assisting device not only to meet workers' requirement for reducing their physical workload but also to retain production flexibility.

A method is presented to display partial slip in precision grip of humans. Partial slip is important information for precision grip of humans, because the human controls grasping force by detecting expansion of the partial slip area. Therefore, a device that displays partial slip will make it easier for remote controlled mechanical hands to grasp objects, irrespective of the friction coefficient between the mechanical hand and object. To design the device displaying partial slip, we analyzed the deformation of a finger when in contact with a rigid plate using a finite element model. We also analyzed the deformation of the finger section model when in contact with the device that displays partial slip. As a result of the analyses, it is confirmed that there is a possibility in displaying partial slip using a tactile display device, which stimulates the finger through deformation of an elastic body.

The study concerns a human error recovery function for efficient human/robot collaborative parts conveyance tasks, which has not been discussed so far in construction of human/robot coexistence systems. We propose a human operational intention inference updater which allows a robot to judge whether the goal intended to reach by the operator is correct or not, from incomplete trajectory data sequence of the robot endtip maneuvered by the operator. Experimental results showed that a use of the proposed inference updater reduced the task execution time even when human error occurred in a task, which demonstrate the usefulness of the updater.

Recent advances in vision technology lead to the establishment of non-verbal communication channels from human behaviors toward understanding their intention. An adaptive vision-based attentive tracker (AVAT) is proposed for isolating such human intended actions from the ordinary walking behavior. The algorithm which drives the attentive tracker is divided into two sub-processes: one is for modeling the movement of human body parts as the environment using hidden Markov models, and the other is for learning the model of the tracker's action using a temporal difference algorithm. In the paper, we describe in detail the integration of the two sub-processes and finally show an experimental result of isolating the human sign action during his natural walking motion for demonstrating the feasibility of our proposal. Identification of the sign context determines the action of the tracker with the simulated rewards supplied.

It has been pointed out that the hard structure of an EVA glove deteriorates efficiency of tasks in the space environment. We also found a claim that an EVA glove did not allow an astronaut to acquire contact information at the fingertips. In the study, we proposed a SkilMate Hand for space EVA gloves which has both a tactile media and a power assist devices. We locate SkilMate in a wider framework of wearable intelligent machines which assist in affording such working surroundings that they can exhibit their skills in spite of their necessity for wearing special suits typically in hazardous environments. The paper focuses on the development of a finger system of a SkilMate Hand, and begins with determination of the design guidelines through analyses from an interview with an astronaut. To improve the task efficiency, we manufacture a power assist device which compensates the bending moment exerted at a human finger joint utilizing an ultrasonic motor. We show the structure and the control strategy of the motor drives. To overcome the second problem of tactile insensibleness, we produce a tactile media device which is composed of a slip sensor element on the outer side and a vibrotactile display element on the inner side of a SkilMate Hand. Transduction characteristics of both pressure sensitive conductive rubber and piezo-rubber are experimentally examined in the low temperature region. Finally, piezo-rubber is chosen to be used as a slip sensing transducer.

An artificial elastic finger skin for robot fingers was developed for controlling the grasp force when the weight and friction coefficient of the grasped object are unknown. The elastic finger skin has ridges at the surface to divide the stick/slip area. It also has a pair of tactile sensors embedded per one ridge similar to human fingertips. The surface of the whole finger is curved so that the reaction force can be distributed. A finite element (FE) model of the elastic finger skin was developed to perform a dynamic contact analysis using the FE method in order to design the elastic finger skin. The elastic finger skin was then constructed. It was confirmed by calculation and experiment that the incipient slippage of the ridge that occurs near the edge of contact area can be detected. This result is useful for controlling the grasping force when the weight and friction coefficient between the elastic finger skin and grasping object are unknown.

We propose a human-will-based approach to constructing a coexistence system in which a robot is assumed to assist the will power of a subject while reflecting his intention and desire. We discuss the model of human's will associated by the work - psychological mind reading, and identify the advantages of the system in which the robot guides the the subject manipulating along the path intended by him. Then, we describe two major techniques with experimental results: 1) estimation of the human-intended path pattern from the previously assigned one by proposing a modified algorithm of hidden Markov model; and 2) merging of the human's force feeling pattern into the desired one by proposing a field impedance equalizer which allows both the human to express his desirable field-dependent force/speed feeling and the robot to display the impedance along the previously estimated path. Finally, we conclude the study with some comments on future works.

This paper mainly concerns a vibrotactile sensor capable of generating impulsive high-frequency signals by which only slipping states of an object can be distinguished. Slip information is so indispensable to guaranteeing the grasp of an object and detecting static friction coefficient that various kinds of slip sensors have been proposed. However, there has been no report on such a slip sensor that slip signals supply real-time critical information to extract only a slipping state from various contact states. We propose a sensor structure with surface ridges in the concavity of which strip-shaped sensor elements are vertically fixed. We analyzed the slip signals of the sensor and showed that the output signal in slipping states was originated in the free vibration mode of the sensor elements. Then, we verified, by using a two-fingered robot hand that the slip signals of the sensor could distinguish the slipping state from other contact, release, or rolling state. Finally, we demonstrate that the slip signal of the sensor output guarantees a slipping state to increase the grasping force of the two opposing robot fingers not to drop the object, by making full use of the impulsiveness of the slip signals.

Autonomous distributed control (ADC), as a promising approach in dealing with complex control problems, is capturing the attention of automation specialists, yet there is still no available platform supporting its industrial application. This paper introduces an ongoing project toward this goal, carried out in our laboratory. In this work, we use agents and an agent society to model an ADC system. We propose a new agent architecture for intelligent machines and a supervisor agent for humans to provide supervision control for agents. To facilitate the coordinating behaviors of agents, an agent communication language called CCL (Cooperative Command Language) is developed, and an object-oriented implementation of the proposed agent system is presented, too. Furthermore, the support system under development is introduced in a simple way. A case study concerning the shape control of a variable-geometry truss is also given to show the availability of the agent system.

In this paper, we discuss a method to achieve safe autonomous robot system coexistence (or Kyozon in Japanese). First, we clarify human pain tolerance and point out that a robot working next to an operator should be covered with a soft material. Thus, we propose a concept and a design method of covering a robot with a viscoelastic material to achieve both impact force attenuation and contact sensitivity, keeping within the human pain tolerance limit. We stress the necessity of a simple robot system from the viewpoint of reliability. We provide a method of sensing contact force without any force sensors by monitoring the direct drive motor current and velocity of the robot. Finally, we covered a two-link arm manipulator with the optimum soft covering material, and applied the developed manipulator system to practical coexistence tasks.

This paper concerns a new tactile sensor with both 3-axis force sensing and slip sensing functions. The first three axis force sensing is achieved with tactile heads, each of which is supported by three pressure sensing elements. The pressure sensing function is achieved with arrays of pressure transducers that measure a change in the contact resistance between a specially treated polyimide film and a resistive substrate. The slip sensing function is achieved through the use of a stress-rate sensor that responds to small-scale variations in a surface profile with roughness. Static and dynamic analyses of the new tactile sensor were made and compared with experimental results. Further experiments were conducted with an algorithm to detect rotational slip.>

An alias free parallel structure for adaptive digital filters (ADF's) is considered. We propose a new cost function for parallel ADF's. The limiting value analysis of system identification using proposed cost function is given in stochastic sense. The cost function is defined in each bin; accordingly, it enables the parallel processing of ADF's.>

Numerical applications frequently contain nested loop structures that process large arrays of data. The execution of these loop structures often produces memory reference patterns that poorly utilize data caches. Limited associativity and cache capacity result in cache conflict misses. Also, non-unit stride access patterns can cause low utilization of cache lines. Data copying has been proposed and investigated in order to reduce cache conflict misses [1][2], but this technique has a high execution overhead since it performs the copy operations entirely in software. We propose a combined hardware and software technique called data relocation and prefetching which eliminates much of the overhead of data copying through the else of special hardware. Furthermore, by relocating the data while performing software prefetching, the overhead of copying the data can be reduced further. Experimental results for data relocation and prefetching are encouraging and show a large improvement in cache performance.

Conventional subband adaptive filters (ADF) using filter banks have shown degradation in performance because of the nonideal nature of filter banks. For this problem, the authors propose an alias free subband structure for adaptive filtering using polyphase frequency sampling filter (FSF) banks. As a preliminary, they make it clear that the conventional polyphase discrete Fourier transform (DFT) bank is equivalent to a class of the FSF bank. Then, they propose a new class of ADF using the FSF banks based on the frequency sampling theorem. As a result, the proposed technique enables subband adaptive filtering without the degradation effect of both the aliasing and cross terms, even if one chooses critical subsampling.>

It is very difficult to place tactile sensors around revolute joints of manipulators, since those joints and surrounding surfaces are not stationary and thus they remain insensitive to contact with obstacles. This interferes with the introduction of sensors for collision avoidance control in manipulators. This paper deals with a tactile sensor system for universal joint sections of manipulators in order to detect collisions with obstacles. Its unique characteristic is the ability to detect, in one sampling period, the positions of more than one collision despite having only one signal line. It is achieved by making use of the resonance phenomenon of the sensor circuit, and all sensor elements can be distinguished from one another by frequency division. The structure of the sensor and the method to detect the label number and position of the sensor elements that are in contact with obstacles are described, followed by analysis of the circuits and experimental results of the sensor system.>

Deals with a multi-sensor system for controlling a two-fingered hand to grasp various objects. The multiple sensors are grasping force sensors, a friction coefficient sensor, and a weight sensor, which are mounted on the hand. Algorithms to recognize objects and to control grasping force of the hand by integrating the detected sensory information are also presented. Experimental results of case studies show that the ability to detect grasping force of wide dynamic range, friction coefficient, and weight allows a hand to grasp many different objects.>