In Japan, the population of person requiring support is increasing. Exercise is important for care prevention. So, local government has carried out gymnastics for care prevention. However its participation rate is low. Therefore, improvement of the motivation to exercise is needed for elderly people. So, we develop input device which operation will be exercise for the preventive care and combine it with tele-robot. We expect tele-robotics contributes to daily life and hobby of the elderly, and then the elderly would improve the motivation to exercise. In this paper, we discuss the problem of previously developed hand ergometer type input device and improvement of the device

View full abstract

In order to accomplish general human behavior navigation, we have proposed and developed a Behavior Navigation System (BNS) using Augmented Reality (AR) technology. The BNS using AR have the AR display system, which superimposes the computer generated image of the expert over the real camera image captured by the cooperator. The expert and the cooperator see the superimposed image. The expert can induce the cooperator to move as the expert moves if the cooperator tries to mimic the CG image of the expert. In order to spread the behavior navigation technology, we have developed a novel easy-to-use/limited-performance BNS using a smartphone. Although the BNS is mainly specialized to the tasks, which can be accomplished by one hand, most smartphone user can use the BNS easily. In this paper, we introduce the BNS using a smartphone.

View full abstract

In this paper, a modeling method is proposed for manikin that can be used for education and training of various surgery. To adjust hardness by each part of human body, the manikin is made from UV curing gels (UV gel). The manikin is formed by laminating sheet of UV gel that is reproduced a cross-section of human body. The sheet is provided with fine dimples for being dripped colored UV curing gels. Therefore it is possible to arrange colored UV gel in a uniform distance. And it was confirmed that it is possible to pile up the particles of the UV gel was colored in method of laminating a sheet of UV gel.

View full abstract

A real-time SPH fluid simulator is useful for constructing a virtual environment containing fluid with a free surface. In order to develop a more efficient simulator, it is better to control the mass of each particles, in which particles near the surface has light weight and make high density; and particles near the bottom is heavy and low density. In this research, we study the stability of such simulations including particles with different mass and the characteristics of a feedback gain. Then, we propose a more stable feedback method considering mass change rate. Finally, we perform a simulation to show the effect of the proposed method.

View full abstract

We developed a Virtual Chameleon that is a wearable system to provide independent fields of view to the eyes in previous work. The Virtual Chameleon artificially enables the user to vary the directions of the visual axes of both eyes to arbitrary directions independently and to perceive each field of view simultaneously. Although humans move both eyes in various styles of eye movements, attitude control of camera platforms of the system were only corresponding to smooth eye movements because a control method of the system depends on only arm movements. It was assumed that visual stimuli induced by various eye movements affect predominance occurred in binocular rivalry. Therefore, in this work, we focused on patterns of eye movements and control method, and implemented additional functions to make users control Virtual Chameleon in various ways by building trackball devices into the system.

View full abstract

We propose a concept of giant experience to extend our bodily size limitation. In this paper, we address requirements including binocular vision and much wider pupirillary distance(PD) to let users to feel surroundings smaller. As a result, it is expected to be felt as if being the giant. As a basic prototype to evaluate it, we report a basic implementation of the binocular vision under much wider PD.

View full abstract

In this paper, we present a method to calculate touch forces in a remote place using low-cost depth camera and present those forces to the user in real-time. Using depth camera's captured data, a 3D virtual scene surface representation is constructed from the remote point of view and used to calculate pushing and stroking forces during user fingers movement. The algorithm computes force vectors for each desired point using point's motion vector and calculating god-points against the scene surface. By using such a method, it is possible to avoid using complex articulated arms and hands in the remote side to measure haptic forces at fingers end points. We show here the process of capturing and constructing the environment, and the method of calculating the interaction forces using depth data. Also we suggest using superimposed virtual hands for natural interaction with the remote environment.

View full abstract

The mass of robots with multi degrees of freedom is inevitably large due to achieve high accuracy or output force. Lightweight of the robots is important factor in order to use in life space. As methods of lightweight for robots several soft actuator s made of rubber material have been developed. In our researches, Inflatable and Flexible Actuator (IFA) has been developed. This actuator is a bag shape made of plastic films. This actuator is very lightweight but the rigidity is not sufficient for utilizing as a life support manipulator. This paper describes several structures to reinforce rigidity for the IFA.

View full abstract

A small sized reduction device using elastohydrodynamic lubrication is proposed for robot and mechatronics. The proposed reduction device could be simply and cheaply manufactured by the conventional machining method and easily downsized because of no requirement of tooth profile. The test equipment was constructed in order to utilize to a characteristic grasp of proposing mechanism and the development design.

View full abstract

The purpose of this study is development of the flow control valve which is small in size, lightweight and can control air flow proportionally. In this study, we thought about a condition that makes flow rate proportional to applied voltage taking the vibration property of orifice plate into consideration. To verify whether the condition is appropriate to flow control valve using particle excitation, a prototype using Langevin type transducer structure that aims at application to proportional valve has been designed by using finite element method and made. And we have evaluated the vibration property and flow rate of the prototype.

View full abstract

We develop wave-shaped tubes to improve the bending performance of the bundled-tube locomotive device. If one of the tubes is pressured, the pressured tube inflates and the device bends. When a conventional tube is inflated, its diameter is increased, which prevents the device from moving in a small pipe. The proposed wave-shaped tube enables the device to bend without the diameter increment. The wave-shaped tube has thick and thin regions. The thin regions can inflate largely, whereas the diameter increment is inhibited by the thick regions. Therefore, the wave-shaped tube inflates only in the axial direction. We produced helix and braid type devices made of the wave-shaped tubes. The experimental data verify that these devices have better bending performance than the devices made of conventional tubes.

View full abstract

Backdrivability is necessary to make robots compliant to contact. Such compliance is helpful to perform cooperative tasks together with human and to protect mechanisms from a collision. We developed robot hands that can generate a large grasping force and a high speed motion. One of them uses worm gear mechanisms. A worm gear cannot be back-driven because of the friction between the tooth surfaces. Therefore, we developed a worm gear mechanism that can change its backdrivability by the reduction of the friction using vibration. In this paper, we describe the differences in the performance of the backdrivability by the vibration directions. We experimentally confirmed that the vibration to the wheel axis direction improved the performance of the backdrivability than the vibration to the screw axis direction.

View full abstract

In this paper, a new type of pneumatic actuator for flexible cable robot is proposed. The proposed actuator is based on the catastrophic structure which mainly consists of an air cylinder as the power source, rubbers for energy storage, a magnet as the trigger for energy release and a weight for impact generation. Through experiments we show that the actuator moves backward under low friction conditions and forward under high friction conditions. We also built the simulation model of the actuator for design optimization in the future.

View full abstract

In this paper, we investigate the relation of dynamic property between actuator tension and contraction velocity of McKibben pneumatic actuators via some experiments. Several robots with McKibben pneumatic actuators are widely applied to rehabilitation. On the other hand, it has not well verified how do their characteristics exhibit stable motions from its simple structure. We examine the relation of output tension and contraction velocity and that of tension and length. From the results, we may conclude that dynamic property depending on velocity has no relation to the length of actuator.

View full abstract

Tube pumps have been used in many applications due to the cleanliness and the convenience. It is basically constructed by some rollers which transfers the fluid in the tube by pressing down it from the outside. However, it is known that the flow pulsation occurs when the outflow side roller removes from the tube. Then, we developed a new tube pump which has non-circular gears. The gears realize the move of the roller to achieve the non-pulsation of the pump. We show an experimental result and explain that the gear based on a ratio of angular velocities of rollers achieves the non-pulsation.

View full abstract

This paper proposes a PWM control method based on visual servoing for precise robot motion control. We use a visual deviation between a target position and an end effector position in order to modulate a pulse width. This method does not require precise calibrations for both robot kinematics and visual sensors. The usefulness of the method is demonstrated through experiments using a robotic arm with 2 DOFs.

View full abstract

Pneumatic rubber actuators have high compliance and move flexibly, making them work in the various fields of welfare and industry. However, conventional pneumatic actuators need heavy and noisy air compressor, expensive air valves and air tubes to be driven. These make pneumatic actuator immobile. We have developed a new gas source using electrolysis/synthesis of water. It is the gas-liquid reversible chemical reaction. This paper proposes a new pneumatic rubber actuator with PEMFC. The PEMFC can be controlled and downsized easily and mounted on FMA which has two air chambers. In the driving experiment, this new actuator works successfully only with electricity control.

View full abstract

We have been developing a control method of power assist cart with differential gear and brakes. The object of this research is a cart which has the drive mechanism which has a deferential gear like a common car. In this paper, the control method for performing power assistance of both going-straight operation and revolution operation in the drive unit which consists of one motor on which a user operates a brake by a handle is discussed.

View full abstract

We propose a torque control strategy of a robot arm using ultrasonic motors. The ultrasonic motor (USM) is an actuator that has advantages such as low speed, high torque (good response time), and high-accuracy positioning without gear. These characteristics are suited for a wide range of robot arms from industrial robots to personal robots. In this report, we show the control strategy that changes the torque by changing phase of two AC voltages that applied the USM. In experiment, one of robot arm's joints becomes back-drivable, or generates any torque by the proposed control strategy

View full abstract

In recent years, Japanese society has been aging, engendering a labor shortage of young workers. Mechanism using a pneumatic cylinder is therefore expected to be useful to perform tasks such as day-to-day work support for elderly people. In this study, we have evaluated characteristics of force control of pneumatic cylinder using predictive functional control (PFC) system. The predictive functional control is one of the model based predictive control (MPC) schemes, which predict the future outputs of actual plant over prediction horizon and compute the control effort over control horizon at every sampling instance. In this paper, PFC control performances of pneumatic cylinder are reported.

View full abstract

In this paper, force control of the servo motor using stochastic resonance is proposed. The servo motor control by a micro controller developed, in this research, consists of an AC servo motor with a ball screw. For estimation of applied force, Reaction Force Observer (RFOB) is employed. In the RFOB, the bandwidth depend on the motor speed. Since the ball screw structure has high viscosity, response force responsibility of the estimation is not suitable for practical industrial application. In this paper, to improve the system responsiveness, servo control system with stochastic resonance is proposed. Finally, experimental results show the effectiveness of the stochastic resonance.

View full abstract

Pneumatic robots have structural softness which is an important element when robots work in living environments of humans. In addition, pneumatics is one of the approaches to realize jumping, running, walking and so on, which are performed by animals and humans. The self-contained pneumatic robot we previously developed has a commercial super-small air compressor, which generates insufficient pressure for dynamic motions such as jumping. To solve this problem we designed and developed super-small high-pressure air compressors. As an additional method of improving the maximum pressure, we tested to increase the intake pressure by using the compressed air stored with a regenerative air pressure system. We verified the efficacy of supercharging. We also developed a self-contained pneumatic robot BIRDLEG with the new super-small high-pressure compressor and the regenerative air pressure system. We verified the validity of the design concept for self-contained pneumatic robots by realization of jumping on BIRDLEG without any external air pressure source.

View full abstract

In Japan of the super aging society, various wearable robot for daily life assistance and rehabilitation care are developed. Those robots almost form an external skeleton along a frame of the human body, but the movement of the wearer is restricted to the flexibility of the rigid robot links in the structure. We developed a non-exoskeletal robot with a little feeling of restriction using skeletal system of the human body, but must support all the weight and the inertial forces of the robot in the human body. In this study, we devise a lightweight motor unit with flatness necessary for a non-exoskeletal robot and produced it experimentally.

View full abstract

This paper evaluates lifetime of Rubberless artificial muscle. We found the cause of short lifetime of the basic type Rubberless artificial muscle is crease of the air bag. Therefore, we tested several improvements in order to avoid creasing the air bag. The lifetime became 100 times as long as a basic type Rubberless artificial muscle by these improvements.

View full abstract

This paper reports the posture control of a spherical joint antagonistic drive system using a rubberless artificial muscle. This system is a robot joint in which the spherical bearing is driven by two antagonistic structure of a rubberless artificial muscle. The angle measurement device attached to spherical bearing measures the posture of joint. Experiment is a posture control of a joint using two DOF PI control system, at the case of a changing the antagonistic force or a changing the joint posture. Results of the experiment verified that this drive system can move with a target value in these cases.

View full abstract

The purpose of this study is to enable a mobile robot to localize itself in a crowded environment. Scan matching is an example of localization method. However, it is not effective in environments with many dynamic obstacles, i.e., crowded environments. In this study, we propose a method to obtain data usable for scan matching in crowded indoor environments by the process of scan data accumulation and near point deletion. With this method, we can obtain wall data that is similar to when there is no crowd. Experiments were carried out to investigate the effectiveness of the method. As a result, the method was found to be effective in localization in environments with many dynamic obstacles.

View full abstract

Accurate self-localization system is necessary for a mobile robot to work in a home environment. Studies of a indoor autonomous mobile robot often use vision sensors and laser range finders. It is difficult to estimate the exact global location if the environment has number of places that have similar shape boundaries or small number of landmarks to localize. Furthermore, it tends to take a long time to recover the self-localization estimation if it goes wrong at once. Self-localization system based on reading RFID-tags on floor is robust against changes of lighting condition, obstacles, furniture and doors conditions in the environment. This paper proposes a self-localization method based on multiple RFID readers and dense IC tag textile and shows its validity in a real environment.

View full abstract

New idea to keep the accuracy of distance measurement in a practical use is proposed. The property that parallel lines in the disparity space are converted to parallel lines in the real world under the accurate disparity offset is used. The experimental result shows that the correct disparity can be obtained within the accuracy of 0.05 pixels.

View full abstract

In this paper, we propose a indoor environment mapping method for large space with multi-robot. We build a large space map with map fusion by the maps that multi-robot built. Our system is composed of 2DLaserSLAM, the mutual recognition with the camera and map fusion with ICP. When moving robots recognized it mutually, we integrate maps formed from each other's 2DLaserSLAM using ICP. The effectiveness of the proposed method is proved through experiments in indoor environments.

View full abstract

In this paper we focus on the problem of Visual Odometry. In recent years, several VO algorithms have been SAD or ZNCC in stereo matching and feature points tracking. These methods are template matching with luminosity distribution and have properties of accuracy of pixel and poor change of lightness of the image. These properties cause low accuracy of correspondence of feature points. In this work we employ Phase-Only Correlation in stereo matching and feature points tracking. POC only use phase of image and have the properties of high accuracy of subpixel and strong in change of lightness. We experiment with image data set of rough terrain and experiment result show that this method provides the high accuracy of VO.

View full abstract

In this paper, we propose a method for travelable area detection using a 3DLIDAR. In order for the autonomous mobile robot to travel safely, travelable area detection is necessary in such areas where traveling becomes hard by differences in the height and material changes. Our system is integration of evaluations suitable for each such factors that make traveling hard. The evaluations are composed of the height information, the curvature estimation, the classification of the reflection intensity. Afterwards, we construct a local environment map of the robot from detected results. The effectiveness of the proposed method is proved through the experiments results in outdoor environments.

View full abstract

In simultaneous localizing and mapping (SLAM), odometry and range sensor are generally used. There is a problem that errors are accumulated in odometry process. To address this problem, information of fluorescent light positions on the ceiling in a room, which is known in advance, is used to determine the absolute position and orientation of the robot. Using absolute fluorescent light information instead of incremental odometry information, the mapping accuracy is expected to be improved. A cleaner robot Roomba was adopted as a mobile robot, on which an omnidirectional camera for detecting the fluorescent lights and a laser range finder for detecting the layout of room are installed. Using this robot, the effectiveness of proposed mapping and localizing method was experimentally proven.

View full abstract

The purpose of this research is to develop the image processing module to support autonomous running of Automatic Guided Vehicle. Automatic Guided Vehicle is a kind of automatic carrier and generally used in distribution warehouses, but it has some problems, such as cost. The Automatic Guided Vehicle using for this research detects the two-dimensional code for the marker on the floor. The developed image processing module acquires the information of the current location and position gap from the picture of two-dimensional code taken by USB camera. The image processing program acquiring the position gap information had been made.

View full abstract

In this paper, we propose an azimuth estimation method for mobile robots. In our method, artificial walls within the detection range of the LIDAR mounted on the robot are used for estimation of the azimuth angle. Each wall is approximated as a plane, and the normal vector of this is used for some processes. By means of this reasonable approximation, the surrounding environment is described with extremely little amount of data. This leads the real-time property of the method. Additionally, taking account of the kinematic performance of the robot, robustness and high-responsiveness is realized. At the end of this paper, the utility of our method is shown through two experiments.

View full abstract

In this study, we aim to achieve an autonomous locomotion of the mobile robot in the unknown indoor environment. In particular, we propose a novel landmark construction and detection method. The landmark is generated by combined image and shape features. In the landmark detection step, we represent the new matching with confidences defined automatically. The effectiveness of the proposed landmark is verified by experiment. Moreover, we introduce the novel landmark based graph SLAM. In our method, that landmark detection is performed on each node. Then, if the robot find the landmark it found once, local loop closure is generated and optimization is performed. The advantage of this method is that we can perform graph optimization before find loop-closure. The effectiveness of this proposal graph SLAM is verified by simulation and experiment.

View full abstract

In this paper, we propose movement locus acquisition of a paddy weeding robot using a simple pixel coordinate transformation method, because of an algorithm that guesses unweeding area from movement locus of the paddy weeding robot. First, we measure position information of the paddy weeding robot by cameras attached overhead and obliquely upward in the experiment filed. Then, the position coordinate obtained from obliquely upward camera are transformed into position coordinate obtained from overhead camera using a simple pixel coordinate transformation method. Experimental results demonstrated that the proposed method is effective for obtaining movement locus of the paddy weeding robot.

View full abstract

When we use odometry for localization of the robot, accumulated error which caused slip of tire become the problem. Currently, by the spread of the internet, map information is easily obtained. Therefor we can obtained priori information such as building and roads without go directly to the place. In this paper we propose the method which reduce the accumulated error by matching outline information of the building that was extracted from digital map and wall surface information of the building that is obtained from LIDAR. Our experiment was performed over 1000 m trajectory approximately in Meiji University to verify its validity.

View full abstract

In this research, we developed a system that can offer information of the location and floor of in the building to using pressure sensor, Gyro sensor and encoder. Pressure sensor is used to get floor number. Encoder and gyro sensor is used to get location by odometry calculation.

View full abstract

We aim to achieve robust vision-based localization for outdoor mobile robots. One of the difficulties in outdoor visual localization is to overcome severe illumination conditions. In strong sunlight, ordinary cameras suffer from under or over exposure because of their limited dynamic range. Recently, cameras having a wide dynamic range (HDR cameras) have been available. In this paper, we evaluate our localization system using an HDR camera. We carried out several experiments: a) how much under or over exposures can be suppressed, b) robustness of detecting and matching SIFT keypoints, c) accuracy and robustness of outdoor localization. We demonstrate that HDR cameras can significantly improve the performance of visual localization under drastic illumination changes.

View full abstract

This paper presents a self-localization method using particle filter for robot with multiple sensors. Multiple sensors do not necessarily work effectively, unless considering environmental dependency of them. To select appropriate sensor set at each time steps, robot needs to evaluate sensors dynamically. We propose a method based on standard deviation of particles and showed the effectiveness of our approach on a computer simulation.

View full abstract

This study addresses the localization problem of the wheel type mobile robot by using the odometry method. The tire slip is most important problem of the mobile robot localization by using the odometry method. In this study, we proposed the odometry method with the Luenberger type observer in order to consider the dynamics for the drive system of mobile robot. The simulation and experimentation results performed in order to evaluate the effectiveness of the proposed method. Simulation and experimental results demonstrated the capability because the localization error by a slip was reduced by adopting the estimation system.

View full abstract

As a first step, an autonomous robot needs to know where is himself in the environment. Simultaneous Localization and Mapping (SLAM) is a method of estimating its own position and creating a map of the surrounding environment in unknown area at the same time. In the conventional SLAM method, a wheel rotation angle (wheel odometry) has been need, but it cause some problems, the method is vulnerable to tilt and slip, difficult to porting to other robots, especially, the robot without wheels. Therefore, we propose the method of SLAM using IMU-based Visual Odometry and 3D-LIDAR. We use 6DoF estimation from a visual odometry instead of wheel odometry input, and to generate one-line laser scan from 3D-LIDAR. As a result, we obtain the real-time SLAM without wheel odometry, and demonstrate that it can apply not only to wheel mobile robot but also to human hand-held.

View full abstract

Range-based ego-motion estimation method, which is so-called "scan-matching", has an important role in robotics. However, conventional scan-matching method, assumes that the error of obtained range data is constant. This degrades estimation results when the robot uses sensors whose error is not constant, e.g.stereo vision. Furthermore, we must also be taken into account for the error in the ego-motion were estimated. In this paper, we propose a scan matching-based 6DOF ego-motion estimation method that considering range ambiguity of the stereo vision and error in the ego-motion were estimated. Simulation experiments are conducted to show the effectiveness of our method.

View full abstract

In this paper, we explore the challenging 1-to-N map matching problem, which exploits a compact description of map data, to improve the scalability of map matching techniques used by various robot vision tasks. We propose a first method explicitly aimed at fast succinct map matching, which consists only of map-matching subtasks. These tasks include offline map matching attempts to find a compact part-based scene model that effectively explains each map using fewer larger parts. The tasks also include an online map matching attempt to efficiently find correspondence between the part-based maps. Our part-based scene modeling approach is unsupervised and uses common pattern discovery (CPD) between the input and known reference maps. This enables a robot to learn a compact map model without human intervention. We also present a practical implementation that uses the state-of-the-art CPD technique of randomized visual phrases (RVP) with a compact bounding box (BB) based part descriptor, which consists of keypoint and descriptor BBs. The results of our challenging map-matching experiments, which use a publicly available radish dataset, show that the proposed approach achieves successful map matching with significant speedup and a compact description of map data that is tens of times more compact. Although this paper focuses on the standard 2D point-set map and the BB-based part representation, we believe our approach is sufficiently general to be applicable to a broad range of map formats, such as the 3D point cloud map, as well as to general bounding volumes and other compact part representations.

View full abstract

Robot self localization is a fundamental problem for autonomous robot navigation. The objective is to take a view image acquired by the robot as query input and to estimate the self pose with respect to view images in a map database. In this paper, we focus on the standard SIFT features based self localization framework and improve its robustness by means of geometric verification of the configuration of SIFT local features between images. We demonstrate the effectiveness of the proposed method via self localization experiments.

View full abstract

Autonomous mobile robots have been required for many service robot systems. In many studies, it is necessary to prepare metrical maps before autonomous navigation of the robots. Although such metrical maps can be generally built by moving the robots and collecting sensor data of fields, it is takes much time to build accurate maps. In this study, mobile robot navigation with topological maps is considered. Especially, we propose a link estimation method while running of a mobile robot on a topological map, using map matching technique utilized for car navigation systems mainly. In this paper, we explain link estimation method and experiments with an actual robot based on proposed method.

View full abstract

Recently robotic localization and mapping have been a field of study. They are the basic function of a mobile robot. However such methods are not enough in the dynamic environment including human and moving objects. In such field, there are serious problems such as occlusion and fewer landmarks, that make localization error. This paper describes a nobel method of localization and multi-layered 2D mapping in the dynamic environment using selective update for particle filter. It is possible to localize precisely, in spite of negative factor in the dynamic environment. Also we can exactly make multi-layered 2D maps.

View full abstract

This paper proposes a new global scan matching algorithm using the CIF descriptors and a geometric constraint between keypoints. The CIF descriptor was proposed in our previous work. It is an feature descriptor that is invariant against a congruence transformation. Our method can perform global scan matching in a cluttered environments without using an initial alignment. Through experiment in real environment, we confirm the validity of our method.

View full abstract

This paper describes a new positioning method which improves the availability and accuracy of multi-GNSS positioning in urban canyon environment using QZSS satellite. In our proposal method, we use GNSSs including GPS, QZSS, GLONASS, Galileo, and Beidou. Furthermore, we use the QZSS as an only master satellite. We conducted a static evaluation test in a narrow-sky environment with many obstacles above the GNSS antenna.The results of the static evaluation test show that our proposal method is highly available and has accurate positioning (fixed ratio: 92.7%, horizontal RMS error: 4.2cm) in urban canyon. In contrast, the number of DD measurements of the conventional multi-GNSS positioning method is 4 or less nearly all the time, indicating that this positioning does not work most of the time.

View full abstract

In vast environments and poor visibility environments, sensor field of view is insufficient and Rao-Blackwellized Particle Filter SLAM often fails to building maps. We propose a new SLAM method using proposal distribution with sliding window matching of accumulated scan data. The proposed method compensates for the lack of field of view using accumulated scan data.

View full abstract