DELTA-4 realizes 3 degrees of freedom of translational motions by a novel redundant parallel link mechanism. These benefits can be obtained by DELTA-4 comparing with conventional parallel mechanisms: wider working area, smaller footprint and easier access to the wide range of working area from operators. In this paper, evaluation experiments of DELTA-4 are described about an accuracy of force display, a mechanical stiffness and a back drivability. In addition, an evaluation experiment of force display is presented.

View full abstract

There are haptic devices as devices that display the virtual reality to human. It is difficult to display the feeling of tactile sense either hardness or softness on hardware in past haptic device and it cannot display both tactile sense. This paper presents the design and development of a modified delta mechanism with variable stiffness. The proposed mechanism can be used as a haptic device to display virtual force and haptics. Since the stiffness at the endpoint of the mechanism is variable, the mechanism can display hardness as well as softness of a virtual object or virtual environment. Preliminary experiments are performed to validate the variable stiffness characteristics of the mechanism.

View full abstract

This paper describes a new tactile device which produces stress fields in 3D space. Combined with mid-air and/or 3D stereoscopic displays, this device provides high-fidelity tactile feedback for interaction with visual objects. The principle is based on a nonlinear phenomenon of ultrasound; acoustic radiation pressure. The fabricated prototype device consists of 324 airborne ultrasound transducers, and the phase and intensity of each transducer are controlled individually. The total output force within the focal region is 1.6 gf. The spatial resolution is 20 mm. The prototype can produce sufficient vibrations up to 1 kHz. An interaction system including the prototype is also introduced which tracks user's hand and provides suitable touch feeling.

View full abstract

We propose a pen-shaped handheld haptic display that allows haptic interactions with virtual environments by generating kinesthetic sensations on the user's fingers; the user's movements are not restricted since the device does not have mechanical linkages. Unlike conventional haptic displays that provide vibrations, which are not representative of tactile sensation, our proposed device is developed on the basis of the hypothesis that the kinesthetic sensations on fingers alone are sufficient to represent the sensations of touch and provides kinesthetic sensations to the muscles in the user's fingers. We verified our proposed method through an experiment with a prototype device and confirmed the representational ability of our device. We also constructed a haptic augmented reality system in that the user can see and touch a computer graphics object.

View full abstract

Force display is an important element to set up a user-friendly man-machine interaction system (e.g. haptic interface system). To display force using a simple mechanism, we focus attention on the static characteristics of an omnidirectional mobile robot. This paper presents development of a prototype of a wheeled motion base consisting of omnidirectional mechanism, the static characteristics of the motion base and the preliminary experiment (current monitor calibration) in preparation for estimating motor torque to adjust the generated force.In addition, we show an force display method and the generated force.

View full abstract

This paper proposes a segmentation method of human manipulation task based on measurement of contact force imposed by a human hand on a grasped object. We define an index measure for segmenting human manipulation task into primitives. The index is calculated from the set of the measured contact forces at all the contact points during a manipulation task. Then, we apply the EM algorithm to the set of the indices in order to segment the manipulation task into primitives. These primitives are mapped onto the robotic hand to impose appropriate contact forces on a grasped object. In the experiments, manipulation tasks performed in daily human life have been successfully segmented.

View full abstract

In this paper, authors proposed a novel multi-legged object transporting system by a herd of autonomous robot agents that can cooperate with human operator. The system is named as "Universal Unipods" ,which consists of a lot of unipod robots to be attached on a single load as a body. A unipod robot is simple in mechanism and it doesn't communicate each other unlike many other multi-agent robot system. The system is designed for not only universal use but also all kinds of users which include aged or handicapped people.

View full abstract

We propose a control method for task assignment of multi-robot cooperation inspired by the adaptability of bacteria, which alters gene expressions when the environment changes. Attractor selection model has been proposed for the adaptability to the environment. We use this model for task assignment to make robots adapt to changing environments. Results of a cleaning task simulation show that the method we proposed can achieve cooperation for multiple robots. Although the method is not so fast, it is simple and robust to different environments.

View full abstract

In this study, we developed technique for cooperative assembly task by plural position-controlled manipulator. The technique achieves high positioning accuracy by using a sensing method named point-face touch. As a result, force control using expensive, breakable sensor is unnecessary for cooperative assembly task. An effectiveness of this technique has.been verified by experiments. This experiment is peg-in-hole task. When space between peg and hole is about 0.1mm, the peg-in-hole task succeeded. There was little contact of peg and hole.

View full abstract

Cooperative manipulation by multiple industrial robots with parallel grippers is achieved using slippage between grippers and the object. Some experiments are done to verify that all states of slippages can be realized by controlling the manipulation forces and the grasping forces applied to the object.

View full abstract

In order to present the contact and collision with environment to human operator arm in a natural manner, a novel haptic interface was developed. The key technology is that we use lightweight components (resulting in wearable design) and antagonistic DC-driven tendons. The latter allows not only generating the required torque but also compensating the excessive torque when presentation of rapidly raising torque is necessary.

View full abstract

The hanger reflex is an involuntary motion of a head when it is sandwiched by a hanger made of wire. We have been investigating this phenomenon and have made an interface that induces the reflex. Using this interface, we investigated the necessary condition of hanger reflex. Contrary to our previous assumption, we found that back temporal region of head may play major role in the phenomenon.

View full abstract

Recently, some power assistance systems are developed to help human daily life. Such systems directly reinforce man's physical strength mainly to assist elder people. Unlike such attitude, we suggest different concept to expand man's physical body. More specifically, we are considering third arm. If we obtain the third arm and control it smoothly, new motions can be generated and impossible tasks get possible. In this paper, we developed interface to control a robot arm as our third arm. This interface is a controller device input by human mouse motion, because mouse is a body part having high degree of freedom next to hands. In the experiment, we controlled the actual robot arm using our developed interface. Finally it was confirmed that our proposed interface works effectively.

View full abstract

This paper presents the design concept and the computational model of the downscaled Haptic Recorder system. Haptic Recorder is a system developed for measurement, recording, visualization, modeling, and reproduction of the manipulation skill in utilizing a tool in reaction to the real environment. The downscaled Haptic Recorder system is designed on aim of limiting the function to the measurement, recording, visualization of the mechanical interaction in the proficient skill. The simplified design of the arm mechanism which is targeted to measure the skill of scalpel operation, and the sensor layout and computation for converting the instrumented data to the visualized 3-D model data is described.

View full abstract

Recently, the advancement of virtual reality technologies can provide various force sensations to users. With regard to studies on the haptic interface, most researchers have focused on the development of the haptic interfaces. So, these devices are constructed by using several decelerators. it is supposed that various types of decelerators individually cause the different effects on in the force/position display performance; it may also affect human force sensation. Then, this study tries to evaluate impressions given by the devices, are displayed virtual stiffness. SD method and factor analysis have been applied to evaluate the impression. As a result, decelerators and control methods affect impressions like activity and smoothness.

View full abstract

In this paper, we discuss the manipulation of planar objects by multiple cooperating mobile robots using the concept of Object Closure. It is the condition under which the object is trapped so that there is no feasible path for the object from the given position to any position that is beyond a specified threshold distance. We proposed the concept of Dynamic Object Closure for achieving object caging task that robots team is able to cage a moving object after a predefined time interval. A path planning algorithm by using Random Caging Formation Testing algorithm (RCFT) is proposed for realizing Dynamic Object Closure. Some simulation results are presented for illustrating the validity of the proposed algorithm.

View full abstract

In this paper high-speed manipulation system is developed, which consists of two manipulators and active visions. The design concept and the features of the system are described. In addition a task using non-contact state is taken up in order to achieve high-speed dynamic manipulation. Experimental results are also shown in which a manipulator throws a ball and the other manipulator hits the ball.

View full abstract

This paper addresses a new cooperation technique for multi-agent system. This technique is based on human social behavior. Multiple robot cooperation has a potential to flexible and adaptable for task variations. In this paper, the personality of each agent is controlled autonomously by interaction between agents. We focus on human behavior imitating other persons who work well. We implement an egoistic parameter and an altruistic parameter as personality to the agents. By the combination of these parameters, the personality is decided. Each agent updates these parameters automatically every constant time to improve performance of the whole system.

View full abstract

Recently, the productive system that recognizes changes in the environment is requested on the production site. There, to recognize changes in the environment we use cameras. Moreover, the system composed of the camera and the manipulator is called a hand-eye system, and the hand-eye calibration is needed in this system. Even if the error from the calibration position of the camera and the manipulator is minute, re-calibration is needed in this calibration method. It takes time to the calculation of the projection matrix to make the image coordinates and the three-dimensional coordinate correspond in re-calibration. Then, the change in the position of the camera and the manipulator is recognized by the landmark in this study. And, we proposes the technique for correcting it to former three-dimensional position without calculating the projection matrix again by correcting the image coordinates, and effectiveness is verified by the experiment.

View full abstract

We do research on an object recognition system for a service robot that can help humans by getting objects by a user's demand. In this paper we propose a scheme to segment and detect individual objects. At first, we divide foreground including objects from backgrounds by filtering based on hue variances in a HSV color space. This filtering also has the effect to remove redundant edges like ones of printed figures on the object surface. After edge detection operation is executed, the image is divided into some regions separated by some edges. Yet, segmentation is not perfect. The ambiguity of regions on the edge still remains. Thus, we propose the robot with an arm touches or taps objects. The visual information changes as a result of physical contacts between the robot and objects. This change makes region separation clear.

View full abstract

We will discus about our new measuring method that can detect EEG under exercising by using an adaptive filter algorithm that we designed. Usually, measuring EEG is done in the quiet state. In case of the measuring EEG under exercising, the gap of the electrode occurs by the vibration of the body by the exercise. It becomes noise and affects the detection of the exact EEG In order to remove a noise, we tried to measure the EEG under exercising used adjustment correlation filter.

View full abstract

The sensor with magnets to detect and trace the light gauge steel is designed. It has detecting mode and tracing mode to keep high position accuracy and high scanning speed. It is installed to the dismantle robot and it is tested for the light gauge steel detecting and tracing. The test showed that the design method is useful.

View full abstract

Because it is non-invasive and easy to measure, EEG that is measured using surface electrodes is used for various objectives. Recently, various methods for BCI (Brain Computer Interface) are developed in a lot of laboratories. Our objective of this research is the development of methods that can be able to establish BCI using EEG that is measured using surface electrodes. In this paper, we will talk about the EEG measurement system for using this objective.

View full abstract

BCI is a direct communication pathway between a human or animal brain and an external device. Our objective of this study is a development of the method that becomes to be able to control various equipment by using EEG. In this paper, we will talk about our developed electrode's construction for BCI.

View full abstract

If a measuring EEG under exercising is possible, we can assess the mental condition of an athlete and the results of this assessment will be useful for mental training. Our objective of this research is a development of the remote sensing method of EEG under exercising. In this report, we evaluate new telemeter system and new electrode about EEG measurement system.

View full abstract

This paper proposes optical measurements of inner diameter and inclination angle of deep hole. The light of an infrared LD is emitted from a slender probe to the surface of an object and the scattered spot light is focused on 2-dimensional PSD (position sensitive detector) through a long Selfoc rod lens. It was confirmed that the accuracy of the measured distance between the probe and the object wall was less than 20 μm.

View full abstract

We have developed an effective non-contact technique for the measurement of the electric potential of the human body when it performs a walking motion. This technique is based on the detection of an electrostatic induced current generated by the walking motion. The electric potential of the subject is determined by integrating the electrostatic induced current flowing through an electrode that is placed at a distance of 1.5 m from the subject with respect to time. The values of the estimated electric potential are qualitatively in good agreement with those obtained in previous experiments by using the ordinary contact-type hand electrode.

View full abstract

In this study, an analog circuit for target tracking system was proposed and designed by using analog circuit for motion detection based on the biological vision system. Analog circuit for detecting sound localization was also proposed based on the biological auditory system previous proposed. The proposed circuits were evaluated by the simulation program with integrated circuit emphasis (SPICE). The simulation results showed that the circuit based on the vision system can output the signal for target tracking and the circuit based on auditory system can output the signal for detecting sound localization.

View full abstract

The demand for crazing inspection robot rises to improve work efficiency of the sewer pipe inspection. Since it cracks in the surface in a pipe and a pipe bonding link etc. is in except, there is a problem which a robot incorrect-recognizes. In this research, it cracks using the discrete Haar-wavelet transform, and a candidate is recognized. Furthermore, the bonding link of a pipe is recognized using the Hough conversion, and how to detect a crack is described.

View full abstract

I research and develop it aiming at the realization of a scooping goldfish robot copied from the method that a human being takes. This paper describes I estimate the 3D position of the goldfish in the water tank based on the information from a camera on a water tank to distinguish a timing to scoop a goldfish.

View full abstract

in case of disaster such as a severe earthquake, many sufferers will be generated at the disaster area. In this situation, we have a problem that sometimes rescuers are also injured by disaster. In order to avoid such damage, various kinds of rescue robots are being developed and used. if some vital signs can be measured by a rescue robot, the efficiency of rescue operation will be greatly improved, in this paper, we will talk about a developed method that can measure pulse and SPO_2 Using this method, these vital signs can be measured only by touching the sufferer's skin with a developed sensor.

View full abstract

This paper discusses non-contact stiffness sensing of elastic tissue with coupling effect by using only displacement measurement during a force impartment. Supposing that both applied force and the point displacement of tissue are known, we can obtain stiffness parameters of living tissue by utilizing a single layered model with horizontal and vertical stiffness units. From both analysis and simulation, we confirm that the parameters are nicely estimated under the condition of at least two different sets of the force and the displacement with appropriately selected initial condition.

View full abstract

To save life of more victims at the stricken area, it is important that rescuers divide a victim by triage. Effective rescue operations can be realized by it. However, well trained person for the triage can do correct judgment, but it is difficult for other people. Therefore, if rescuers gather bio-signals of the victim when they find or save the victim, they can get a clear criterion. From this, it is thought that many rescuers can evaluate the condition of the victim correctly by using bio-signals of the victim. So, we developed the device to detect an electrocardiogram (ECG) by easy contact in this study. Then, we adopted the measurement system by capacitive coupling to be able to use at the stricken area. In this paper, we will describe developed method that can measure ECG by easy way.

View full abstract

There are approximately 300,000 visually impaired persons in Japan. Most of them are old and cannot make appropriate use of their white cane. Therefore, it is imperative to develop some instruments that support the independent activities of the visually impaired. Our objective of this study is a development of a simple and inexpensive navigation system for the visually impaired which can use in the indoor space. In this system, a color sensor installed on a white cane senses the colored guideline, and by vibration, the system informs the visually impaired that he/she is walking along the guideline. RFID tags and a receiver for these tags are used in the map information system. An antenna for RFID tags and the RFID tag receiver are also installed on a white cane. The receiver receives tag information and notifies map information to the user by mp3 formatted pre-recorded voice.

View full abstract

The system that recognizes hand motions using SEMG is experimentally developing. In this system, a 96 multi-channel electrode is set on forearm and 96 channels SEMG are measured. The recognition rate is usually up to 95% by using optimal 4 channels of them and canonical discriminant analysis. However, neural network is usually used in same objective. So, in this paper, canonical discriminant analysis and neural network are compared in the recognition method.

View full abstract

We propose a navigation system for the visually impaired using GPS. First, we prepared a GPSreceiver and a computer. The next, we made programs to navigate safely using location data, digitalmap and voice data. These programs searches for the most safely route and inform about a route tothe destination using voice data. We tested this system and installed on a cell phone.

View full abstract

SEMG (Surface EMG) is one of the bio-electric signals that is generated from the muscle. There are many kind of muscles in a human body, and used muscles are different for each body movement. Therefore, we can detect the body motion form generated SEMG patterns. We use 4channel SEMG signals to detect right hand motions. And our objective is the control of a radio-controlled toy by using the results of hand motion detecting. Five basic hand movements are recognized using canonical discriminant analysis. Two normal subjects were tested for our control system. For every subjects the recognition rates were up to 90% and they were able to control radio-controlled toy very well.

View full abstract

This paper introduced a high-speed vision platform, JDP Express, which can record image sequences of 512×512 pixel at 2000fps as well as real-time image processing. IDP Express was also applied to a high-speed active vision system for target tracking to capture image sequences at high resolution, which were magnified by tracking a target in an image. Several experiments were actually executed for automatic high-frame-rate video recording by using image-based trigger, and the videos captured at a high frame rate were also shown to verify its effectiveness.

View full abstract

In this paper, we propose a method to recognize periodic gestures from images as a key technology for human machine interface. The proposed method uses power spectrum and phase spectrum that are obtained by applying FFT to time series of intensity images. FFT is applied to each pixel of the low-resolution images. The method consists of 2 steps. First, the method detects periodic motion regions from power spectrum. Second, the system uses the phase spectrum in the detected periodic motion region to classify the gestures. The proposed method is robust to lighting condition and individual difference of skin color, because it does not use color information. To demonstrate the effectiveness of the proposed method, we have conducted several experiments and show that support vector machines are well suited for obtaining discriminate hyperplanes for recognition of periodic motion.

View full abstract

This paper proposes an object detection method using characteristics of spectral reflectance in mid- infrared region. Retroreflective traffic signs are irradiated with infrared light and the reflected light from the traffic signs is taken with an infrared camera. Infrared lights with wavelength of around 1400nm and 1150nm are used, since solar lights do not contain these wavelengths and the detection of traffic signs can be realized without the effect of the sun. The effectiveness of the proposed method is verified by experiments.

View full abstract

With the advance of processing ability of processors, more autonomy is expected for various kinds of robots. To enhance the autonomy of robots, various sensors are employed on robots. Especially, we can derive various information such as speed, distance, object, material from vision sensor. However, to extract these information from vision sensor and realize realtime image processing, vision sensor should be more intelligent and have pre-processing ability on itself. We employed a silicon retina sensor as vision sensor and add higher level functions using FPGA. In this paper, we explain velocity estimation of mobile robot by using optical flow obtained from silicon retina.

View full abstract

In the recent years, motion capture systems are used for robot vision and man-machine interface. Most of those systems need attached markers on human body and several sensors. So makerless motion capture system is needed. In this paper, we propose the technique for posture estimation from monocular image sequences. This technique estimates a posture based on detecting body parts. Body parts (arms, legs, head, torso and so on) are detected without markers using human body structure model. A posture is estimated by constructing detected parts. This technique does not require a manual operation to determine initial posture at the first frame. A markerless motion capture system is realized by using this technique. In the experiment, we confirmed realizability of this technique.

View full abstract

We developed an acceleration visualization marker using moire fringe which can display a stripes pattern that corre spond to a magnitude of acceleration. The proposed marker consists of two elastic plates and two plates printed with strips and needs not electric elements such as an amplifier and strain gauges. When proposed marker takes a picture by using a high-speed camera, an acceleration can be measured from remote place. We have experimentally verified the utility of the proposed marker by taking a picture with the camera, and processing the image.

View full abstract

A wide dynamic range image sensor with a real-time image processing capability was implemented. The sensor comprised of FPGA circuits and a silicon retina, whose resistive networks enable instant spatial processing including smoothing and edge enhancement. In order to widen its dynamic range, the sensor exploited the technique in which the voltage signal of the photo sensor was read out several times during a single accumulation time. The multiple images with different sensitivity were integrated in the FPGA circuits. The implemented system was successful in taking edge-enhanced wide dynamic range images.

View full abstract

We have developed a tracking system to provide a Cartesian robot with shape and pose of the human hand, in order to grip or move an object in an assembly task. Particle filter is applied to estimate hand shape and pose with two kinds of hand models. One hand model is a palm model for controlling a robot axis, and the other is a sideward model for gripping or releasing an object. We carried out two kinds of experiments for hand estimation. Experimental results show that the proposed system can estimate hand position approximately and recognize the difference between motions for gripping and releasing.

View full abstract

This paper describes a pose estimation method using a 3D textured model of object that a robot has targeted for handling. The proposed method estimates the pose of this object using the POSIT algorithm and matching SIFT features generated from both the camera image and projected images of the model from a range of postures. Also, we achieved fast computation of this method using Graphics Processing Unit.

View full abstract

When a robot manipulates some objects, the object should be specified by the robot vision. This paper describes a novel object recognition approach which combines a method for generic visual categorization with three- dimensional segmentation method. The approach is as follows; first, the object area in the image is cut out by clustering 3D map of features obtained by SLAM. Next, by using the generic visual categorization method, "Bag-of-Keypoints," to the area, robot can recognize what it is and where it is in the image. Our approach contributes to the object recognition that can correspond to the diversity of objects and the image segmentation that uses 3D positions of features which compose the object. In this paper, we discuss the effectiveness of our approach through some tableware detection experiments.

View full abstract

In this paper, we propose the method of human gait estimation using a wearable camera. The proposed method has an advantage in achieving the estimation regardless of surroundings. Human gait is estimated based on Extended Kalman Filter by using some human walking models which are defined as example motion clips, and an observed image which is captured by the camera attached on human leg. The experiment shows that this method allowed the accurate estimation even when an user walked at different velocity.

View full abstract

In this paper, we propose an image recognition technique for picking robots. The proposed technique is based on the seed fill algorithm which can separate an image area into each object, and the rotation invariant image matching (RIM) which is a template matching. Since the RIM scheme uses the distance information between the center of gravity and the edge pixels of the object image, it is not influenced by changes in posture of objects.

View full abstract