Long-time interaction between user interface and in-vehicle information system often brings about traffic accident. To solve interface design issue reducing driver distraction inducted by the interaction, a new interaction interface is proposed in this paper. The two key features of the interface are: It provides multi-modal display, tactile mode and visual mode. The tactile channel to recognize the screen information releases other heavily loaded sensory channel and reduces reaction time. Next input pattern (sequence) and surface shape are adaptive to user. According to usage frequency, surface shape is deformed and proposed to user so that user can recognize it and can do input operation in easy and quick manner. In addition input patterns are modeled depending on user's situation such as sex, age, personal biorhythm. So it offers user optimal input pattern as soon as possible than routine input pattern while reducing navigation error. To confirm the effect of keypads, experiments were performed comparing with touch screen providing routine surface shape.

View full abstract

This papar presents a software interface that enables users to control home robots remotely. To control robots easily and safely, we propose users using a remote control interface based on Mixed Reality. The software interface displays the simulation how robots will move and requires confirmation of their operations. The simulation is one of the easiest ways for users to acquire the result from their operation, and confirmation will prevent robots from acting out of users' expectation. This interface is also useful for developers for robots. It visualizes the sensors which are attached to the robots and it also visualizes status of robots. When the developers utilize these data, it will help them to understand the meaning of these data. We implemented the software interface and demonstrated it. It enabled us to control home robots easily and safaly.

View full abstract

Even though it is plane images, people are able to perceive a sense of perspective from them when polarized light glasses or head mounted displays are used. However, using such devices impose a burden to people since people should wear them and they only can be used in restricted space. In this paper, 3D display method based on UD-1 and Anamorphose theory is proposed to solve the above problems.

View full abstract

We investigate on development of myoelectric elbow prostheses with ultrasonic motor. Ultrasonic motors have some special unique features, such as high-torque under low-speed operation, compact size, and no electromagnetic noise. On other hand, use of the threshold level or amplitude of EMG (electromyogram) is often adopted as control methods of practical myoelectric prostheses using microcomputer. However, this method is not always suitable for the proposed prostheses. In order to development of an EMG controller for the myoelectric elbow prostheses, we proposed an EMG control system using PSoC microcomputer for ultrasonic motor. In this study, we development the EMG control system for ultrasonic motor using PSoC microcomputer, and it was proved effective in fundamental experiments.

View full abstract

In this study, we try to develop the cloths type assist device to support those human's standing up motion, which is one of necessary and frequent movements in our daily life. The developed assist device consists of pneumatic rubber artificial muscle. It makes easy to wear in daily life to improve the QOL. In this paper, after describing a structure of the assist device, the effectiveness for assist of standing motion is verified through some experiments.

View full abstract

We developed a power assist machine attached to the upper limb, using differential gears for nurses who care an old person being bed ridden, or workers who often lift loads up and down. Two motors behind the shoulder of the equipped person provide motive power to drive shoulder and elbow joints, and the differential gear makes it possible to divide the power between them effectively. This differential gear system can reduce the backlash of gears. The five degrees of freedom of shoulder blade motion were taken into account with the shoulder-joint link mechanism. Furthermore, to grasp the behavior of the arms of the equipped person, pressure sensors were attached under the upper and lower arms. As a result of the experiment wearing the machine, the arms of equipped person could fluctuate at will.

View full abstract

Several rehabilitation robots for upper limbs have been proposed so far, and clinical effectiveness was reported in several studies for the aged people or patients with stroke. However most of them have only 2-DOF for its active motion. It is important to design a rehabilitation system which trains in the 3-DOF space because the upper limbs of human work in 3-DOF space even expect for the wrist. We developed the quasi 3-DOF rehabilitation system which has 2-DOF force-feedback function in working plane but its working plane can be adjusted the inclination. And we named it Hybrid-PLEMO for it can be switched between active type and passive type. Hybrid-PLEMO is a compact, low-cost rehabilitation system for upper limbs with high safety by using ER brakes or ER actuators. In this paper, we describe the mechanism of Hybrid-PLEMO.

View full abstract

The objective of the study described in this paper is the development of a 5 degree of freedom (DOF) externally powered transhumeral prosthesis for above elbow (AE) amputees to increase their mobility in daily life activities. The proposed prosthesis is supposed to be controlled by using a combination of the electromyogram (EMG) based control and extended physiological proprioception (EPP) based control. A multilayer artificial neural network to classify the daily activities using the arm joint kinematics is used to design the EPP based controller. The arm joint kinematics was acquired with a VICON motion capture system while performing the various daily activities.

View full abstract

Authors have been developing an electric prehention orthosis for people suffering from paralysis on their fingers. One feature of this orthosis is that the orthotic part to be put on user's hand is extremely compact and light in weight. This is because the actuation unit is separated from the orthotic part. Two strings in flexible tubes are used to connect the actuation unit to the orthotic part and work as a driving force transmission mechanism. The other feature is that the grip force can be continuously adjusted by users anytime as they want. The user's intention of finger operation is detected by an original sensor that can detect muscle activity over clothes. Based on the user's intention an electric motor in the control unit is actuated thus user can operate his/her fingers with well controlled operation force. The detail of the prototyped system and results obtained from some evaluation tests are reported in this paper.

View full abstract

In this paper, we have developed a master-slave robot hand. Because of the safety issues raised by robot contact with people, we developed a robot hand that uses balloon-type pneumatic actuator with these issues in mind. The master-slave robot hands works flexibly and imitate human's hand movements by using sensor glove. But since we use ON-OFF pneumatic valves, it is difficult for an artificial muscular type actuator to control high accuracy. So we made the joint model of the five-fingered robot hand, the control system of the joint model has been designed and experimented to confirm the effectiveness of the control system. The result shows that PI control work well to use in robot hand effectively.

View full abstract

We propose a practical robotic follower to support daily life. The robot carries a burden and follows a person with a certain distance like a porter. Our ultimate goal is to develop and commercialize a practical robotic tool as a consumer product which can be used in outdoor environment. In this paper, we focus on the "Hyper-Tether" concept for robust sensing for the foregoing person and discuss its application to carry an oxygen tank for a patient with home oxygen therapy.

View full abstract

Guide-Dog Robot which includes moving mechanism, detection and avoidance of obstacles, and human-machine interface functions is being developed for a visual impairment person. A stereo method is verified as a device to detect a pole-shape obstacle and expand to a bar-shape obstacle under in-house condition. Robot can estimate a coordinate of the obstacle using the information by the stereo method. To avoid them and go to a temporary target without any land mark, we propose two kinds of potential function and an additional potential of a person. Then, we can show arrival to the target without any troubles by using prototype robot.

View full abstract

A three-wheeled omnidirectional cane robot is designed for aiding the elderly walking. Possible walking modes are analyzed and a corresponding hybrid model is constructed to describe walking behavior. A concept called intentional direction is presented to denote the moving intention of a human. Based on experiments and some assumptions, dynamic model of intentional direction is obtained as well as its online estimation method. A new admittance control scheme is presented based on intentional direction, which provides natural and intuitive human machine interface. Experiment results show the effectiveness of the design.

View full abstract

There exist handicapped who need help with transfer from sitting posture to sitting posture. They want to transfer without helper, when one feels like doing. Many of transfer assist tools are not the equipment that help self transfer of handicapped, but the equipment for the reduction of helper's load. In previous works, we have proposed a self transfer assist robot that the handicapped can transfer to a desired position without helper. In order to increase usefulness of the robot, in this investigation, we develop a mechanism for mobile function of the robot, which is compact and powerful.

View full abstract

In this study, we propose a leg-robot with an MR clutch to realize virtual haptic control for spastic movements of brain-injured patients. This system can be used in the practical training for trainees of physical therapy. Additionally, we will study to figure out the physiological mechanism of spastic movements of human with the process to simulate patient-like spastic motion by this robot. In this paper, basic structure and mechanism of the leg-robot with the MR clutch are explained. Finally, experimental results of some kinds of haptic control for spastic movements are described.

View full abstract

In this paper, we propose a patient robot whose feeling changes as varying time by making use of chaos orbit to create human's psychology and we develop a system for the effective injection training. For confirming the effectiveness of this patient robot, according to experiment with patient robot using by student nurses, we measure heartbeat of student nurses at four different experiment conditions and we get questionnaire concerning with injection training and patient robot. In addition, we propose visual servoing by making the patient robot do some motions like human. Visual servoing to 3D head pose has been performed by the motions of eyes, neck and waist of the patient robot to keep the human's front face in the center of input images in left and right eyes.

View full abstract

Development of fabrication technology made it feasible to fabricate micro order structures. Those technologies include machining, MEMS (Micro Electro Mechanical System) and laser fabrication. In micro-world, assembly becomes especially a crucial problem. We have developed an easy and fast assembly process of microparts called STAMP (Stacked Microassembly Process). In this process, three-dimensional structural object is fabricated by stacking up two-dimensional layers and micro-structures are cut apart after all the assembly is completed. Here, as an example, we have fabricated a model of stag beetle. Two-dimensional layer parts are fabricated by SU-8 photolithography and electroplating. Stacked parts are bolted together using micro-rivets.

View full abstract

Endoscopical tool to manipulate and peel tumor skin in ESD (Endoscopic Submucosal Dissection) is proposed. This microarm will come out from the tip of an endoscope to help the manipulation of the tumor skin. Inexpensive fabrication process is a must to make the tool disposable for sanitary reason. The arm is fabricated by photolithography and electroplating and assembled by stacking up the layer of electroplated parts (STAMP: Stacked Microassembly Process), thus mass-production with low cost is possible. The key element of this microarm is that it has a joint. Measuring the angle of the joint can calculate the tip position of the microarm, thus being able to control the arm. An angle sensor will be placed in the future. The actuation of the arm is done by guide wire.

View full abstract

This paper reports fabrication of a new micro-device for biopsy using Stacked Microassembly Process (STAMP). By using STAMP, we can manufacture a three-dimensional structural object. This biopsy device has barbed micro-spikes and an opening mechanism. We fabricated parts for the biopsy device by Ni electroplating and assembled the biopsy device by using STAMP. A shape memory alloy wire was fixed under the needle to actuate the opening mechanism. We confirmed motion of the opening mechanism by controlling the SMA wire. We also performed an experiment of tissue sampling test. From the fluorescence images of the sampled cells around the tip, an opening mechanism with barbed micro-spikes was effective for biopsy.

View full abstract

To support disabled persons with less muscle strength like muscular dystrophy patients, we are developing simple manipulator systems. In this paper, we report the interface part of the system. One-click button is introduced to input method for the manipulator control. The input menu panel is designed to perform various actions of the robot with the one click button. A manipulator simulation system is constructed to evaluate the input performance. All components are constructed with the RT middleware.

View full abstract

Various input devices are developed to operate a service robot. These input devices can be connected to a service robot systematically by constructing the interface module as RT-Component. This paper describes a user interface component for manual operating a service robot using OpenRTM-aist.

View full abstract

In this research, "Pretty" and "Loveliness", etc. is investigated, and bipedal walking type pet robots with which the senior citizen can be familiar is developed. In this research, the mechanism for the communication behavior in addition to the walking operation was examined.

View full abstract

This report proposes a hardness detection sensor that aims to evaluate the hardness of biomedical tissue. The sensor is composed of a hard cylinder material, a soft column material, and two thin plates and strain gauges. The forces which act on the soft and hard materials are measured by strain gauges through a deformation of the thin plates. The ratio of the forces evaluates Young's modulus of a soft contact object.

View full abstract

This paper discusses a stiffness sensing system that can measure tongue stiffness in vivo. Considering that the global stiffness generally depends upon the thickness of tissue, we utilize the equivalent Young's modulus defined by the applied pressure divided by the tissue strain. We also introduce an index of muscle contraction by using EMG signal for judging whether the stiffness increase is caused by a muscle contraction or a disease. By using an experimental system composed of a gripper typed stiffness sensing system and an EMG sensor, we show that the equivalent tongue Young's modulus under muscle contraction is four times larger than that under no contraction.

View full abstract

This paper discusses a concept of a noninvasive sensing method that can estimate a left ventricular stiffness towards a medical diagnosis. Focusing on not only the strain of ventricular wall but also the displacements of epicardium during diastole of heart beat, we introduce an index of ventricular stiffness based on a two-layer elastic model. Applying the proposed method to the echocardiography, we show an example where we can estimate the difference between soft and hard ventricular walls.

View full abstract

This paper presents the development of an active tactile sensor using fluid. Measurement of stiffness and surface conditions, such as sliminess, smoothness, and others, of soft objects is discussed. This sensor is made of a silicone balloon catheter. The balloon is contacted with a soft object and expanded by using fluid. In the process of the expansion, the balloon pushes the object and the contact surface of the balloon slips or sticks. These phenomena depend on stiffness and surface conditions of the object. The expansion can be evaluated by measuring the flow of the fluid. The balloon that is the sensing element is very soft and flexible and does not need electricity. This sensor is available as a medical sensor since it provides safety for human. Experimental results on various measurement conditions support the validity of the proposed sensor.

View full abstract

The measurement of ground reaction forces is important in gait analysis. This paper proposes a wearable sensor system that measures ground reaction forces by using flexible tactile sensors. First, the design concept of the sensor system is discussed. Second, the method to extend the measuring range of the flexible tactile sensor is explained. Third, on the basis of the concept, the sensor system is developed. Finally, the dynamic response of the sensor system is experimentally examined.

View full abstract

Robot systems are gradually becoming more common in order to perform tasks that people can't. A multiple robot system is very difficult to work because the environment is very uncertain and information is often incomplete. Collision avoidance is a key problem, as autonomous robots must not collide with anything. There is pressure to use less and cheaper, lower quality sensors for widely spreading. Multiple robot systems will use collision avoidance algorithms which are based on low quality data, to in support of this situation. In this research, we designed a collision avoidance algorithm which use reinforcement learning to allow the robots to learn the most effective strategy. Computer simulation experiments were conducted in varying settings. We also experimented with real robots: autonomous lawn mowers RL500, and similarly good results were obtained.

View full abstract

In this paper the running environment was evaluated for the route generation under the real environment with the mowing robots. The mowing robots (Robomower) will be expected to mowing work at open areas like as parks. At open are as like as parks robots should be small and low in power. Because, large and strong robots work at open areas like as parks, then they may injure some people. The mowing robots strongly receives the influence from the running environment because it is Small and low in power. It is thought that the influence is the running resistance between the lawn and the mowing robots. The running resistance in the lawn is not so considered. In this paper investigated route generation that considered the direction of the lawn and running resistance.

View full abstract

This study is to propose a configuration of a mobile robot system that is able to achieve a new movement under the situation where some of its actuators are broken and replaced by alternative ones, which may not be the same configuration as the original ones. The proposed method is based on a Reinforcement Learning and is modified so that it can achieve rapid conversion over a wide search space. To this end, a "growth of action-value" method is proposed, which enables effective exploration of an action space based on temporal reliability of each action-value. A series of 3D simulation-based experiments are conducted, where the proposed method shows rapid conversion to a good candidate of movement patterns.

View full abstract

The purpose of this paper is to develop a simulator for evolution of building structural objects by virtual creatures. As a first step, we develop a simulator of a virtual creature which builds a structural object. We use Open Dynamics Engine to develop the simulator. The virtual creature evolves its way to build structural objects with Genetic Algorithm. We carried out two experiments, building protective barriers and building nests to capture quarries, to indicate the possibility of evolution of building structural objects. As a result, we slightly show evolutionary tendency in structural objects built by the virtual creature.

View full abstract

It is necessary for a mobile robot to obtain a path, which can adapt to unknown and changeable environment. A method of path search using reinforcement learning is proposed. The problem herein is that the path is not necessarily the shortest only by applying reinforcement learning. On the other hand, A* algorithm is one of the method which can search the shortest path. However, in A* algorithm, the position of the obstacle, the start, and the goal must be known in advance of the searching process. To overcome these problems, in the present paper, a method is proposed in which A* algorithm is combined with reinforcement learning. The effectiveness of this method is investigated by computer simulation.

View full abstract

This study deals with development of simulator for the distributed robot control system. In the distributed robot control system, active parts of a robot arm such as motors and links are modeled as software components which make performance according to states of other components based on exchanging messages. This paper reports the development of the simulator for evaluating the distributed robot control system and estimating possibilities for actual robot arm control system, and case studies of redundant robot arm applications.

View full abstract

Reinforcement Learning (RL) is one of promissing approaches for controlling an autonomous robot. However, its performance is quite sensitive to the segmentation of state and action spaces. This paper describes an RL mobile robot, the task of which is passing through a narrow short route instead of a wide but long route. The robot needs appropriately segmented state and action spaces to avoid punishments, otherwise the robot tends to fail to pass through the narrow route. In order to overcome this unwanted situation, we apply our proposed technique, named Bayesian-discrimination-function-based Reinforcement Learning (BRL). We investigate the performance of BRL through computer simulations and analyze the learning process.

View full abstract

This paper presents a neuro-controller for a two-link manipulator with an unactuated first joint, wherein the motion of the system is confined to a horizontal plane. The control objective is to move the first joint from a given position to a target angle. Radial basis function neural networks (RBFNs) is used as the neuro-controller, and particle swarm optimization (PSO) is adopted for the learning algorithm. The effectiveness of the proposed approach is verified by a comparison of numerical results and experimental ones.

View full abstract

We have been developing a reinforcement learning technique called Bayesian-discrimination-function-based Reinforcement Learning (BRL). This is recognized as an effective technique for autonomous specialization, which is a new concept for cooperative multi-robot systems. Homogeneous robots with BRL learn behavior to be heterogeneous so as to play different roles for developing cooperative behavior. In this research, BRL is applied to arm-type autonomous robots, the task of which is lifting an object without tilting it. Then, we investigate the autonomous specialization process by means of observing behavior transitions about their role assignment.

View full abstract

This paper presents a control method of swarm agents using Boids model. The swarm agents are considered as uncontrollable targets, and they are controlled by a small number of user-control agents with introducing communication to specific agent to the swarm agent. Simulation results show that the proposed control method works well when the swarm is in a steady state.

View full abstract

The exploring mechanism and firming mechanism for effective behavior are revealed in the SONE from our visualization and analysis of its learning process. These visualization and analysis were tried against the process of the online real-time structural reinforcement learning in a mobile robot that learns collision avoidance.

View full abstract

In this paper, we propose a sequential learning model that can generate behaviors for successfully performing various tasks. The model generates the actions based on change in state pattern. The model updates the memorized relationships between changes in sensory information and a motor command through sequential learning. We confirmed the performance of the model by applying it to a mobile robot simulation. The results indicate that suitable behaviors for all the tasks emerged by the sequential learning.

View full abstract

Learning dynamic multi-DOF motion has been a challenging task for real robots. With past learning algorithm, the number of required trials has been too large for real robots. And the use of simulation suffered from dichotomy between the simulated world and the real world. In this paper, we report on a series of learning experiments where a real robot acquires dynamic whole body action skills from scratch, solely based on real trials. We use our original learning algorithm named "Exploration of Virtual Goal Switching Pattern" which discovers semi-optimal solutions extremely quickly. In order to cope with deterioration of the search performance due to the real world noise, a new improvement to the search method is devised. A series of experiments with real robot is carried out in which the robot automatically achieves the initial posture and repeats the trials. The results show that dynamic rising actions are automatically acquired from scratch, including a variety of strategies.

View full abstract

In this study, we apply neural network-based balancing control to a 1-legged hopping robot with a spring element and seek continuous robot hopping under adequate balancing. We first built a remote control system for our robot that was composed of a hopping mechanism with 1-DOF and a joint mechanism with 2-DOF and created teacher signals for the neural network using a joystick-like interface. Then we embedded the obtained controllers into the robot and evaluated the performance for balancing control. As a result, we finally achieved six continuous hops; at the same time, we clarified the difficulty novices experience when they attempt to remotely keep our hopping robot in a balancing state. For supervised learning, the robot's behavior basically depends on the quality of the teacher signal. Our experiments also suggest that the enhancement of the method to generate teacher signals is crucial for more sophisticated neural network-based balancing control.

View full abstract

This study, I accurately modeled underwater environmental influence and am aimed for modeling of the behavior of the object in the environment. As a result of these modeling, I report that it is possible that the behavior simulation of the object in the underwater environment.

View full abstract

The control and mechanical systems of an embodied agent should be tightly coupled so that useful functionalities such as adaptivity can emerge. This indicates that the mechanical system as well as the control system should be capable of a certain level of "computation" for generating the behavior. In order to explicitly indicate this type of "embodied" computation to be embedded in the mechanical system, Pfeifer et al. have recently coined the term called "morphological computation", which is expected to be an indispensable concept for building adaptive agents in various time scale such as here-and-now, ontogenetic, and evolutionary time scales. In this study, we focus on ontogenetic time scale of "morphological computation", and intensively discuss how the change of some body properties through growth influence stable learning to achieve bipedal walking. The preliminary simulation results derived indicate that the change of body parameter through the growth from child to adult plays crucial role in achieving efficient dynamic walking.

View full abstract

This study aims at acquirements of an object (animated robot) motion under the air environment based on physics modeling. We introduce several resistance forces of the air into physics modeling as the air environment. Simulations on falling objects and gliding one in the air verify that their movement reflects with the realistic world phenomenon in the air.

View full abstract

Recently, it has been widely recognized that the mechanical system as well as the control system should be responsible for a certain amount of computation for generating the behavior. However, there still leaves much to be understood about to what extent "computational offloading" from the control system to the mechanical system should be achieved. In order to effectively consider this concept, this study focuses on long-distant physical interaction between body segments as a characteristic physical property of the mechanical system, and discuss its effect from the viewpoint of learning. To this end, we have developed a two-dimensional serpentine robot with monoarticular and biarticular muscles which effectively induce long-distant physical interaction between body segments as a practical example.

View full abstract

This study proposes Anibot (Animated Robot), which can behave autonomously in the virtual environment. A difference between Anibot and a conventional animation is that Anibot behavior follows a physical law. This report focuses on modeling tools newly developed to create Anibot, and simulates learning and controlling for an elastic model of Anibot.

View full abstract

In this paper we tackle a Fish-Catching task under a visual feedback hand-eye robotic system with a catching net. For the sake of such innate ability being widely existed in animal behavior, the catching operation becomes tough and some effective intelligent method needs to be conceived to go beyond the fish intelligence. We embed chaos and ramdom motion into the net motion to realize a kind of robotic intelligence, and we shown the chaotic and random net motion is effective to overcome the fish escaping strategies. The effectiveness of the chaotic and random motion is confirmed through saccessive fish catching experiment.

View full abstract

In the 3D-CG animation, many creators waste their time on creating the motion of object. Therefore, we develop a new modeling software for rigid objects. This software has the automatic modeling function based on the lows of physics. This paper describes the structure of modeling software, and shows examples of creating motions.

View full abstract

This study focuses on emergence of heterogeneous personal characteristics in a society. Such characteristics can be observed in not only human but also many organism, and deeply effect on personal making-decision and action. This paper treats with a swarm-agent model in order to examine how and what heterogeneous personal characteristics become diversified. A pursuit problem is set for the swarm-agent to efficiently capture a run-away object in cooperation. First, numerical simulations which the number of agents is changed, we observed that agents' characteristics become diversified. Next, complex network is adopted for the swarm-agent communication and we observed swarm-agents' behavior.

View full abstract

In this paper, for purpose extending applied area of humanoid robot, we study recognition method for a changing motion in the water and guess the depth of the water. The property of the action in the water is different compared to on the land compared that because the influence of resistance inertia force from the wave. Therefore, the action control is more difficult. We use a small humanoid robot "Robovie-M" wear the diving suit.

View full abstract