In this study, a power assist wear has been developed. The device is wearable like a normal clothes. It is lightweight and designed in a way to blend with everyday apparel. In this paper, we discuss the development of the power assist glove. The glove is the first development in the series of power assist wears.

View full abstract

We developed a slave manipulator and its controller for microsurgery in various surgical fields. This slave manipulator targets neurosurgery and vitreoretinal surgery thus far. On each arm, there are three translational motion axes, two rotational axes and three additional axes for detachable instruments depending on the type of surgery. Detachable axes are used to drive the forceps. We designed a neurosurgery forceps and a control unit for vitreoretinal as well. Robot controlling software was developed on LINUX with real-time extension. This software composes of communications module, trajectory calculation task and servo controlling task. Communications module receives the target data from master manipulator through the UDP network socket. Trajectory calculation task calculates the ideal trajectory polynomial by interpolation of desired targets using the weighted least square method. Servo controlling task calculates the target position and velocity in every servo cycle, and outputs command signal to each servo amplifier.

View full abstract

This paper shows global reconfiguration capability evaluation for redundant manipulator tracking target trajectory and avoiding obstacle control. In this research, one approach for an automation system composed of redundant manipulator and movable camera on-line being able to operate for unknown objects without any preparations is proposed. In a such system, the configuration of the robot should always be prepared to keep the highest avoidance manipulability to evade the object quickly that may appears suddenly in the moving camera view. We evaluated configurations of manipulator by using the index representing Avoidance Manipulability Ellipsoid Shape Index with Potential (AMSIP).

View full abstract

The purpose of this study is to develop a power assist device that can hold the user's arm to reduce burden on furl workers. This device consists of a linkage mechanism in order to use a ratchet mechanism. The ratchet mechanism is used to assist the upper limbs without using an actuator. As a result of simplified structure, it didn't cost much. In this papel the ratchet mechanism is designed to hold the shoulder joint and the elbow joint. The effectiveness of the proposed holding mechanism is verified bv the result of the preliminary experiments.

View full abstract

The purpose of this study is to develop a sensor system for the weeding robot in a rice field. This sensor system is composed of the capacitive touch sensor to detect the rice plants. The weeding robot that we use in this study can autonomously move across the rice plants. The function of the sensor system that is mounted on the weeding robot can detect rice plants and cannot be affected by water. In order to elucidate the fundamental characteristics of the capacitive touch sensor, we carried out experiments with the rice plants. The effectiveness of capacitive touch sensor is verified by the experiments.

View full abstract

The objective of this research is to create a dynamical model for bipedal locomotion like human. We define contact between land surface and foot as constraint and consider that human's walking generally uses surface-contacting of foot and point-contacting of tiptoe or heel. Therefore, it seems that using these constraints properly is important for efficient walking. In this paper, in order to depict these constraints, new model is discussed by changing 4 models based on dynamics through Newton-Euler method. Then, we applied the proposed model to bipedal walking control of leg of humanoid robot and evaluated effectiveness of the proposed model by numerical simulation.

View full abstract

Progress in multimedia technology started from the telecommunication of audio information using telephones. And the spread of the interne can telecommunicate visual information using visual telephones. In addition, in recent years, we realize the sense of touch using a device called as a haptic interface. This technology is expected to play an active role in a wide variety of fields for technology of producing sense of touch in addition to technology of producing sense of both sight and hearing which is used in many applications. However, compared to sense of both sight and hearing, sense of touch has not been assessed quantitatively. This paper clarifies a how to recognize objects in virtual reality using haptic interfaces controlled by impedance control system. Results show that objects' hardness and grasp approach direction are great factors when we recognize and memorize the objects.

View full abstract

With development of multimedia technology, the telecommunication technology such as mobile phone and Internet has pervaded our life. In addition, advancement of virtual reality technologies can provide various force sensations to users. It is expected that further studies of interactive force telecommunication including force display would expand application fields of multimedia services. In this paper, we used force display devices, which are called as 'Haptic Interface'. Though a usual Haptic Interface is only used as the force display device, our laboratory realized a force-feeling telecommunication system by use of two Haptic Interfaces, whose control algorithm uses force and position control Hybrid control algorithm. By use of this system, teaching performances of a line, which can be realized transmitting force feeling from a teacher to a student, are examined under several experimental conditions. These results show interesting human cognition characteristics.

View full abstract

In this paper, we propose a new method that can estimate contact force and finger angle for touch panel devices by measuring transmitted light on fingernail. When we touch a light source with our finger, the light (mainly red component) can be seen at the fingernail since the red light can go through the finger tissue. Based on this characteristic, we can estimate the manipulatory force by detecting the light intensity at the fingernail because the intensity of the transmitted light is proportional to the applied force. We can also estimate the relative direction of the finger to the touch panel device. Even though the transmission through the finger tissue is diffusive one, we can know the movement of the light source below the finger with camera attached onto the fingernail. Therefore, a transient pattern of the light source is useful for estimating the relative direction between the finger and the device.

View full abstract

Magnetic-field sensitive elastomer (MSE) is a compound of the magnetic materials and a non-magnetic elastomer. This elastomer has a feature in which the viscoelasticity can be adjusted with the external magnetic field. In this study a haptic controllable device for the sole feet with the MSE have been development. First, The material tests and the sensory evaluations of MSE is experimented. In the material tests of MSE, the difference of the elastic variation according to the thickness and the efficiency of haptics in the sensory evaluation is researched. Finally, as a device to control the magnetic field the electromagnet is designed. For the future work, the virtual simulation of the haptics application during the walking motion in different ground will be made.

View full abstract

In this paper, a haptic device is proposed for micro anatomy realization system. Micro anatomy realization system performs brain anatomy in a virtual environment. In this system, two haptic devices for left and right hand are used to operate two virtual autopsy instrument. Haptic device was developed to achieve the operation of instrument. The feature of this device is as follows. To operate close to devices, 7 DOF link mechanism is used as mechanism of device. To present stiff Objects, admittance control is used as force control method of device. From the experimental results, work space and generation force of this device was confirmed.

View full abstract

In the previous study, we proposed an ungrounded pen-shaped haptic display that provides forces on the user's fingers without the use of mechanical linkages and realizes interactions with virtual environments. The device had three motors inside and was able to provided forces with 3 degree-of-freedom (DOF). In this time, for more simplicity and usefulness, we design and implement the device that has only one motor and provide force with 1 DOF. The user can start using the device with less effort than the previous device. We also construct a haptic interaction system with the device in that the user can feel the sensation of pushing 3D virtual objects.

View full abstract

The sheet metal inspectors in automobile industries know that the knitted work gloves help the perception of surface undulation with micro order. The knitted work gloves have some touch enhancing mechanism. We developed a 3 by 3 tactile display based on its mechanism. In this paper, presenting system of surface figure with micro order using the tactile display is proposed. Micro order curvature is virtually presented on a plane/curved surface.

View full abstract

In this paper, we present a new wearable input technique for small computing devices like smart phones. Our input system does not require any visual assistance. We expand an interface of a handheld device to our clothes. We use a built--in camera of the device to detect a position information pattern printed on the clothes. Once the device identifies its position on the user's clothes, it processes an allotted function to the position. A user can utilize haptic and somatosensory feedback in place of visual feedback because our system requires direct contact between the device and the clothes.

View full abstract

Haptic feedback system is required to provide reaction force resulting in the robot hand to an operator. In this paper, we develop a force feedback device to present reaction force in the operator when the robot hand grasping an object. The force feedback device presents reaction force in distal segment of thumb and middle finger, and basipodite of middle finger. The device is very compact, lightweight and has many force display points by using a Shape Memory Arroy as an actuator.

View full abstract

We developed a master slave teleoperation system of robot manipulators. This system has force feedback function. The master part of this system has two haptic devices and the slave part has a manipulator composed a robot arm and a two-fingered robot hand. This system can control both a robot arm and a robot hand by using mode changes, so operators can control the manipulator by using only their two fingers of one hand. We experimented in two methods. One is a spring device experiment. As a result, operators could distinguish softness of objects. The other is peg-in-hole task experiment. As a result, operability was improved for force feedbacks.

View full abstract

This article proposes a two stage actuator system that consists of a coarse actuator driven by a ball screw and a fine actuator driven by a voice coil motor and a bilateral control system using the proposed actuator system. In the proposed method, by consolidate inexpensive two actuators based on a control design for bandwidth separation, a wide bandwidth bilateral control system can be constructed. It was proved by simulation that a higher transparency is achieved using the proposed system compared with narrow-bandwidth case using a single ball screw system. Furthermore, a prototype of the two stage actuator has been developed and basic performance was evaluated by experiment. The experimental results showed that a mechanical impedance with a mass of 10 g and damping of 1.6 N/(m/s) was successfully implemented.

View full abstract

In recent years, parallel mechanisms have been widely introduced to haptic device for advantageous benefits such as a high rigidity, a high output force, a high accuracy and a high backdrivability by its multi-legged structure and fixed actuators on the base. On the other hand, redundant parallel structure is known that its characteristic can be effectively introduced to avoid singularity and to widen working area. In this paper, an improvement method of forward kinematic model of redundant parallel structure is introduced for further improvement of force producibility. The proposed method can be potentially utilized on other redundant parallel structures.

View full abstract

In this paper, we perform 3D shape reconstruction in the inside of a building by making use of 3D range information from a LRF and image from a camera equipped with a mobile robot. Abstractly, the mobile robot equip with a LRF mounted on a 2 D.O.F. (the pitch axis and the yawing axis) stand to get 3D scan data. Images from the camera are mapped on the corresponding 3D scan data by using a technique of texture mapping. Experimental results in a real environment by using a proposing method are shown. As a result, the robot moves and takes a picture from two or more points, and the mapping from these images to one 3D scanning data is achieved.

View full abstract

This paper presents a new methodology of attitude estimation for cost-effective and small-size inertial measurement units, consisting of tri-axial gyroscope sensors, accelerometers and geo-magnetometers. Introduced for the algorithm development is a quaternion feedback structure, where bias terms in the gyroscope rate information, appearing in the phase-locked loop, are identified with the theoretical guarantee of the global and asymptotical stability. The bias-term identification and modification process is performed continuously without the influence of the disturbance by combining the proposed disturbance detection scheme and a conventional vector matching method. Practical validity of the presented framework is evaluated by experiments. The evaluation result shows that the approach in this paper can offer superior drift-free performance with property of disturbance attenuation and rejection under arbitrary high-frequency test motion of spatial rotation and translation.

View full abstract

In this paper, we propose a method for improving a processing time of a 3D mapping. 3D map is generated from range images measured from multiple viewpoints by registration and integration of multiple range images. A range image is obtained from a 100Hz real-time textured range image sensing system. To improve the processing time of the previous method, Kanade-Lucas-Tomasi feature tracker (KLT) method and Graphics Processing Unit (GPU) for parallelizing some processes is used. The effectiveness of the proposed method is verified to compare with the previous method in experiments.

View full abstract

In car driving support systems and mobile robots, it is important to understand three-dimensional environment widely at once. In this paper, we use a fish-eye camera as a sensor which measures environment. This camera can take a wide-range and distortional image and can be easily mounted on cars. We propose a method for reconstructing three-dimensional environment from fish-eye images based on EPI analysis. Using this method, matching of feature points is easy and stable. The effectiveness of the proposed method is verified by experiments.

View full abstract

This paper describes a three-dimensional environment recognition method for 3-D mobile robot that can walk up and down stairs. The recognition method using conical scanning distance measurement can measure environment shape, location, and posture. The procedures are as follows. 1) Setting up the direction of the scanning area. 2) Measuring distance that follow the conical generating line. 3) The extraction of the planar feature quantity. In this report, the experimental device was built using the TOF range imaging sensor. The measuring experiments about a wall (one plane) and stairs (two planes), and the evaluation of recognition accuracy were carried out.

View full abstract

Accelerometers have direction gaps between the package and the actual sensor in general. Conventional calibration methods of accelerometer, e.g. ISO standard, do not include this direction gap. This study proposes a new calibration method. This method uses newly proposed examination system with parallel link exciter mechanism. The mechanism rotates a test accelerometer. With this exciter, the posture of the accelerometer is kept constant against external coordinates. The input acceleration is stable because required torque of the mechanism is constant by counter weight. From the accelerometer output, a plane including actual sensor direction is detected by maximum value of the sine curve profile. The direction of actual sensor is the line of inter section of two measured planes. Actual sensor directions and sensitivities are experimentally examined.

View full abstract

In this paper, we propose developed three-dimensional distance measurement device for human posture estimation in indoor environment. Our developed device is composed of a horizontal rotary mechanism and a vertical two-dimensional plane distance measurement sensor with mirrors. In addition, our developed device can scan a person fast and densely by tracking human location wherever a person is located at any room place. Because of these advantages, our developed device can measure human shape robustly. In experimental result, our developed device can measure three-dimensional human surface model during tracking a person.

View full abstract

This paper describes methods of environmental map representation and plane estimation for daily assistive robots. It is assumed that 3D information of around environment is constructed by collecting dataset gradually measured, we propose an effective map representation named TCCM (Time-series Composite Cuboid Map). Through several experiments about daily assistance, we ensured the effectiveness of our method.

View full abstract

In urban search arid rescue situations, a 3D mapping technology by rescue robots is very useful to plan rescue missions. Furthermore, it is important to obtain temperature information for discovering victims in disaster scenes. Therefore, we set our research object to construct a 3D thermography map by mobile robots. To realize the above, we fuse both temperature image acquired from infrared camera and range data acquired from 3D SOKUIKI sensor. In addition, to improve a quality of the 3D thermography map, we propose improved ICP(Iterative Closest Point) scan matching algorithm using the temperature information applied to 3D scan points. Through initial experiments, we validated an advantage of our algorithm over traditional ICP scan matching. In this paper, we explain construction method of 3D thermography map, scan matching method using temperature information, and report an initial experiments.

View full abstract

It is a significant challenge to detect or model a transparent object using conventional visual sensors. Here, we show that the characteristics of transparent objects under different lighting conditions (normal light and laser light) can be exploited for improved detection accuracy. Specifically, we analyze the object characteristics and combine them into detection parameters. We propose a method to specify candidate areas on the basis of highlight spots in the color image and then use these areas as bounding rectangle inputs for GrabCut. Thereafter, segmentation by GrabCut is applied to depth images and laser reflectance intensity images. Next, the transparency of the object is determined qualitatively by comparison of the segmentation results for the depth image with those for the reflectance intensity image. Bottle-shaped transparent objects are correctly detected within 1 m of the visual sensor using the proposed method. Evaluation was carried out for different shapes, materials, distances, and backgrounds.

View full abstract

This paper presents a laser based pedestrian tracking by multi-mobile robots in outdoor environments. Each robot finds pedestrians from its own laser scan image based on an occupancy-grid-based method, and it tracks the detected pedestrians via Kalman filtering and Global-nearest-neighbor-based data association. The tracking data are broadcasted to multi-robots through intercommunication, and they are blended based on Covariance Intersection. For the pedestrian tracking, each robot identifies its own posture based on RTK-GPS and laser scan matching. By our tracking method, all the robots share the tracking data with each other, so that they can always recognize pedestrians that are invisible to any robots. The experimental results of tracking three pedestrians by three mobile robots validate the proposed method.

View full abstract

This paper presents a laser-based pedestrian tracking by multi-mobile robots in indoor environments. Each robot finds pedestrians based on an occupancy-grid-based method and tracks them by Kalman filter and Global-nearest-neighbor-based data association. It also generates a local map by EKF-SLAM. When the robots are located in vicinity, they exchange the information of tracked pedestrians and local maps through intercommunication. The tracking data are blended based on Covariance Intersection, and a global map is built by fusing the local maps so that the tracking performance can be dramatically improved. The experimental result of tracking two pedestrians by three mobile robots validates the proposed method.

View full abstract

This paper presents a people tracker with a two-layered laser range sensor(LRS) and a fisheye camera set in an environment. The LRS detects the heads and knees of people; the camera takes color images of people detected by the LRS. From the position data of people taken by the LRS and the color histogram by the camera, heuristic-rule-based and global-nearest-neighbor based data association identify people. Their identified people are tracked via model-based tracker; the interacting-multi-model estimator and multi-model particle filter are applied to tracking people with sudden changes, such as walking, running, going/stopping suddenly, turning suddenly and jumping. Experimental result validates our people tracking method.

View full abstract

We have proposed a new three dimensional environment recognition method using conical scanning distance measurement for 3-D mobile robot that can walk up and down the stairs. This report describes the method to recognize the stair by applying the conical scanning method. The procedures are as follows. 1) Recognizing the positions and postures of multiple regions on 3D-TOF-camera. 2) Linking neighboring regions by using image processing. 3) Extracting the step based on height information. 4) Calculating the position, direction and width of the stair. We devised the algorithm, implemented, and experiments were carried out.

View full abstract

The problem of memory-based robot self-localization is approached from a novel 3D reconstruction technique in the paper. A single-view 3D reconstruction technique employing on the "Manhattan world" assumption is introduced for synthesizing views at various possible viewpoints. Experiments are conducted on indoor mobile robots with an omni-directional vision sensor.

View full abstract

Simultaneous localization and mapping (SLAM) is a fundamental problem in mobile robotics. Map matching approach to the SLAM problem has received much attention in recent years. A main contribution of this paper is proposal of "3D polestar descriptor" to facilitate the fast 3D map matching SLAM. A discriminative shape descriptor and a hash-based indexing technique is presented for 3D point cloud data. Experiments are conducted in indoor environments using Bumblebee stereo camera.

View full abstract

It is very important for robot to find objects and to estimate their poses, especially for a home service robot which works in human living environment. To achieve that, Nakada et al. proposed SIFT-Cloud-Model(SCM)111, that consists of SIFT features, their 3D positions and eye vector to estimate 6D pose of a target. In this paper, we propose the SIFT-Cloud-Model generation method and evaluate its accuracy, arid computational cost. Finally We developed the efficient way to create SCM and cut down about 60 % computational cost of pose estimation method.

View full abstract

This paper describes a visual stabilization with a compact robot vision system that consistsof two CMOS cameras, gyroscope, Field Programmable Gate Array(FPGA), and micro controller. Each camera has two servo motors for pan and tilt rotation. the servos are controlled so that twe target position is kept at the center of the visual field of each camera.

View full abstract

This paper presents a sliding mode filter that effectively removes impulsive noise and high-frequency noise with producing much smaller phase lag than linear filters. It is less prone to overshoot than previous sliding mode filters and it does not produce chattering. In addition, it is computationally inexpensive, and thus suitable for realtime applications. The proposed sliding mode filter employs a quadratic surface as its sliding surface, which is designed so that the output converges to the input in finite time if the input value is constant. Its algorithm is derived by using the backward Euler discretization, which prevents chattering. The effectiveness of the filter was shown by experiments by using an ultrasonic sensor and an optical encoder.

View full abstract

Currentry, many types of no-reversible compressed sound source, represented by MP3(MPEG Audio Layer-3) are popular in the world and they are widely used to make the music file size smaller. The sound data created in this way has less information as compared to pre-compressed data. The objective of this study is by analyzing EEG to determine if people can recognize such difference as differences in sound.

View full abstract

Non-contact stiffness sensor is often utilized especially for medical fields due to its advantage of avoiding damage to tissues and keeping sanitary. However, it is hard to measure the force accurately, because the fluid jet based force much depends upon the shape after deformation of object. This paper proposes an innovative approach where the external force is adaptively calibrated by using C-map where the relationship between the compensation coefficient of the force and the tangential angle of the deformation is experimentally-mapped.

View full abstract

This paper proposes a micro displacement visualization sticker using a moire fringe which can display characters and fringe that correspond to a magnitude of displacement without the use of electronic elements, such as amplifiers, strain gauges and wires. The sticker is simply attached to a measurement object, and it shows displacement information by characters which can be visible to the naked eye. Accurate numerical value of the displacement can be obtained from an image of the sticker by using image processing. The structure of the sticker is simple and its fabrication is inexpensive. The results show that the sticker provides the displacement information by characters and accurate numerical value of the displacement using image processing.

View full abstract

We propose a new proximity sensor for identification of objects. The proposed sensor integrates measurements of permittivity and optical reflectance by impedance measured using electrodes and phototransistors. Using this sensor, objects with different material properties (permittivity, reflectance) can be identified. This technique may be applicable to proximity sensor for robot and factory automation.

View full abstract

We discuss the relationship between normal tension glaucoma (NTG) and high myopia from a viewpoint of the softness of ocular tissue, based on our hypothesis that the geometrical shapes of eyes sufferd by these diseases are compliant and easily expand compared with those of normal. We define a new method to classify the shape of eyes by using geometrical parameters. In addition, we discuss the stress of cornea pressed outerward by internal ocular pressure. We show an interesting tendency where patients sufferd by these diseases have eyes with stronger ellipsoidal shape with statistic significance.

View full abstract

EMG measured the skin surface is widely used as an information source to determine various types of body movements. And there are many research results that use forearm EMG to detect hand motions. In this field, detected results are used in order to control some equipments like as an artificial hand. However most of them only can detect hand motions and cannot detect muscle power for each motion. In our EMG system, we suppose that muscle power is proportional to the amplitude of EMU The power is estimated and the speed of a slot car is able to be controlled using estimated muscle power in our system. Many subjects were studied with our developed system, and most of them could control the speed of a slot car well. In future, we would like to develop an EMG controlled robot hand that can feed back the grip power of a hand.

View full abstract

SEMG (Surface EMG) is one of the bio-electric signals that are generated from the muscle. There are many kind of muscle in a human body, and used muscles are different for each body movement. Therefore, any kind of body motions can be detected by analyzing generated SEMG patterns. Our objectives of this study are the development of a method that makes perfect detection of the hand motions possible using SEMG patterns. And applying this developed method to the man-machine interface. In this study, our target machine that is controlled by SEMG is a robot arm that is made for industrial use. The system analyzes user's SEMG and discriminate 8 hand motions. Then the system controls a robot arm using discriminating results. This system was Studied with some subjects, and most of them able to control a robot.

View full abstract

I can distinguish hand motions by analyzing generated Electromyogram (EMG). A position deriving Surface EMG (SEMG) and identification rule to distinguish becomes important to distinguish hand motions using SEMG. About a position to derive, The good result is gotten by using 96 multi-channel electrodes developed from our precedence research. There is a report when I can cope with the problems of various pattern recognition by using a support vector machine (SVM) about the identification rule, too. These, we develop the recognition system of operation which used 96 multi-channel electrodes and SVM.

View full abstract

EMG measured the skin surface is widely used as an information source to determine various types of body movements And there are many research results use forearm EMG to detect hand motions. In this filed, detected results are used in order to control some equipments like as an artifical hand. However most of them only can detect hand motions and cannot detect musucle power for each motion. If the power that is generated from musucle is able to be estimated by measured EMG, estimated results will be useful to control equipments. In this power, we talk about our developed method that can estimate muscle power of pronation of hand and movement outside times of hand.

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

Development of a BMI(Brain Machine Interface) system using EEG to control machine by thinking. This machine can get EEG easily to put electrode on scalp. Our objective of this research is the development of methods that can be able to establish BMI using EEG that is measured using surface electrodes. In this paper, we will take about the EEG measurement system for using this objective.

View full abstract

In this paper, we describe about a developed navigation system that supports the independent walking of the visually impaired in the indoor space. This system is composed of colored navigation lines, RFID tags and an intelligent white cane. In our system, some colored marking tapes are set on along the walking route. Theses lines are called navigation line. And also RFID tags are set on this line at each landmark point. The intelligent white cane can sense a color of navigation line and receive tag information. By vibration of white cane, the system informs the visually impaired that he/she is walking along the navigation line. At the land mark point, the system also notifies area information to him/her by pre-recorded voice.

View full abstract

We propose touch-game application as a new function of Step-on Interface. It is development of the application for recreations. Users operate the screen projected by the projector by touching by hand. We considered projecting a screen on a desk so that elderly people could perform a recreation safely. This paper shows development of the application of the recreation that elderly people can do it happily and safely all the time.

View full abstract