It is need a base station robot wired a robotic fish to operate the robotic fish by remote control. In addition, the base station robot needs a high maneuverability to response a robotic fish’s action. In the present study, we developed a small robotic fish equipped with Wi-Fi module to send images. Images captured by a CMOS camera transmitted to the operator while the robotic fish swims near the water surface. However, images were not transmitted to the operator in the deep water because the Wi-Fi signals were not reach the receiver. Accordingly, we manufactured a base station robot enable to operate the robotic fish by remote control in the deep water.

We have developed oscillation wing propulsion type fish robots which were modeled tuna .In this paper, first, we modified the defect parts of TDUNA-2013c such as structure of the loosen ligaments, shapes of the dorsal and anal fins, inner space of the body, structure of the caudal fin, battery and connecter, gum boot of the vibration output part, etc. A computer controller was re-installed in the robot to protect robot from out of control issue caused by the receive error of the radio wave. After these modification, we tested the robot’s performance and found that the robot showed good maneuverability about pitch and roll direction. The maneuverability about yaw direction is much worse than the other direction and we found that we had to modify the yaw control system such as the shape and size of the dorsal and anal fin, attachment position, etc.

This paper describes adhesive characteristics of foot pads for insect-inspired wall-climbing robots. Ants can walk on the smooth vertical surfaces using an adhesive organ and secreted viscous liquid. We fabricated three kinds of adhesive pads inspired by the adhesive organ of ants, a glass pad, a PDMS pad and a glass pad with a water supply mechanism. Pure water were provided to the pads, and normal and tangential adhesive forces between the pad and a glass surface were measured. The experimental results show that the hydrophilization treated glass pad generates higher adhesive force than the untreated pad, and the pad with water supply mechanism generates long-lasting adhesive force.

By performing sound source estimation using location information of home robot and a result of sound source localization by HARK, the robot can recognize environmental sounds based on the sound source. Thus, it is possible to distinguish even if similar sounds of feature quantities are generated, so that precision of state recognition of house chores is increased. In this paper, we developed the state recognition system and performed experiments of operation works of a microwave oven. As a result, the localization results were facing a direction of the microwave oven, and so it can be said that it is possible to recognize the sound source accurately.

A customer wants a home robot to work a number of chores. In the previous study, a robot achieved chores by executing sequentially. If chores have a job of a home appliance, a robot must wait to finish it. This wait is repeated when a robot executes the chores sequentially. As a result, a behavior of a robot become inefficient. A robot must work efficiently by executing chores parallel. Therefore, we develop a software system to rebuild the chores to one of a new chore. The system can shorten working time of our supposing chores.

The path planning algorithms for cleaning robots reported so far focused on how the robots move the entire space without waste and missing part. However, considering actual environments, the dirt distribution is not uniform and has some specific features which depend on the shape of the environment and human behavior. Therefore, it is more efficient to clean the region according to the deviation of the dirt. In this paper, we construct a novel path planning algorithm for the cleaning robots which takes the deviation of the dirt into consideration to prioritize the region with large quantity of the dirt and sorts out the parts of the region.

Task knowledge transfer is carried out in various situations such as a cooking training. To actually show how to do a task is important in a task knowledge transfer and includes interactions between an instructor and a learner. Such interactions are divided into three cases: (1) the instructor provides a piece of task knowledge; (2) the learner responds to an explanation of the instructor; (3) the learner asks a question, and the instructor answers it; In this report, we describe an observation system with four RGB-D cameras and headsets for interactive teaching actions. The system detects and records the actions and the interaction between the instructor and learner such as pointing and grasping by using color, shape and human joint position information from the RGB-D cameras. In addition, the system recognizes and records the words that the instructor and learner utter from headsets.

Transfer of task knowledge is ubiquitous in our daily life, and various types of interaction occur in this transfer. Such an interactive task knowledge transfer,however,requires that an instructor and a learner to be at the same place and time. If we use a robot for mediating between them, such limitations could be eliminated. This paper focus on its latter part, namely, human-to-robot task knowledge transfer, in which a robot instructor interactively teaches a human learner how to achieve a task. This has little been dealt with in the robotics domain. We are especially interested in how interactions occur and develop an ambiguity-driven formulation of interactive task knowledge transfer based on the Dempster-Shafer theory. We develop an experimental system for blocks world tasks as a proof-of-concept and show preliminary results.

This research aims to develop a robot which can attend a person.We model the transition of person's state using a finite state machine (FSM). The robot recognizes events for the state transitions and selects an appropriate attending action for each state.We implemented attending actions for the person's walking and the sitting action.When a person is walking, the robot takes a following action.When the person is sitting, the robot moves to a waiting position determined by considering the comfort of the person and the others.We carried out attending experiments using a real robot to show the effectivenss of the proposed approach.

Sphere-shaped mobile robot is expected to be useful as a mobile platform of the home agent robot. The authors have proposed and are studying on a type of sphere-shaped mobile robots using an omni-directional three-wheel internal drive unit. Previously, the dynamics model has been derived and vibration suppression control of the internal drive unit has been achieved. However, velocity control of the spherical shell has not been realized yet. Therefore, in this study, to realize velocity control of the spherical shell, I-P control which uses estimated value of the angular velocity of the spherical shell is proposed and investigated. Simulation and experimental results show the effectiveness of the I-P control.

In recent years, robots having various functions such as cleaning and flying have been developed. It is necessary that such robots estimate the state of people for assistance in daily life. Therefore, We proposed the system in which the robot passes a drink to people who need a drink. The system recognizes people and drinks by the upper camera and the omnidirectional camera. The robot communicates people who don't have a drink on the basis of the recognition result. We conducted the experiments on the line trace and the recognition of drink. As a result, the experiment on the line trace was successful in all cases, and the experiment on the recognition of drink was successful excluding the case when a person has put a pencil case at an angle.

The interface robot was able to perform only a similar reply for various people until now. Therefore, in this paper, we introduce a method for response control of an interface robot suitable for every attribute. We distinguish an attribute based on walking information that get using a laser range finder. It was confirmed by experiment that the robot can respond to each attribute. The attribute expresses a child, an adult, an elderly person .We set up a laser range finder at the height of the ankle. Then, we walked the parallel position away from 1000mm. In the result, the robot was able to response for an adult correctly in the face-to-face.

In this paper, we studied on hand-gesture-based interface for a home mobile robot, using the LeapMotion sensor. The LeapMotion sensore is small device that can sense hand movements in the air above it. We developed simulation programs of operating a mobile robot using the LeapMotion, based on two types of hand-gesture-based interface, ”interface A” and ”interface B”. The interface A adopts ”swipe motion”, which is considered to be an intuitive hand gesture for operating a mobile robot. On the other hand, the interface B uses simple finger shapes. We performed experiments using the simulation programs and compared the two interfaces.

We aim to realize a novel pen based interface system that provides various digital operations for both low-tech and high-tech people. In this paper, we propose a novel sensor fusion method between IMUs (Inertial Measurement Unit) and strain gauges embedded on a pen device. By using strain gauges for obtaining correct direction of pen tip movement during writing, cumulative errors caused by the intrinsic noise/drift of IMU can be corrected. Therefore our proposed sensor fusion method can reconstruct longer handwritten trajectories such as multiple characters including multiple strokes that have not been achieved by conventional methods. This paper provides a pen tip direction estimation method using strain gauges attached on ink cylinder, trajectory reconstruction method using IMUs, and the sensor fusion method of both estimation streams. The algorithms are evaluated by writing experiments conducted by robotic manipulator and human hand.

In this paper, we developed an articulated mobile robot for cleaning. The robot consists of links which has a cleaning unit, active wheel, a pitch joint, touch sensors for collision detection and range sensor for avoid wheels fall off, and the links are serially connected by a yaw joint. We also proposed a control method of the robot to clean a desk without falling off, and carried out simulations and experiments to demonstrate the effectiveness of it. Furthermore, experiments were performed to test obstacle avoidance using touch sensor.

We have been developing a communication robot “KoRo” located at the entrance of the university library. It greets the persons passing in front of it, and accepts hand gestures to contol the display situated beside the robot in order to introduce the library to them. It is also equipped with some games by using visual and sound sensors for attracting the visitors to it. Although the system is still under development, we introduce the framework of KoRo in this paper.

In this paper, we propose an emotion model including mood congruence effect.

The goal of our research is comforting a human by a robot action. A robot action which comforts a human should differ according to the human mood. It seems to be important to estimate and predict the human mood accurately to comfort the human by a robot action. In this paper, we performed action selection experiments with Mof-mof robot which has soft visual aspect. A subject played quiz games and the robot selected an action which was expected to comfort the subject. The action was selected based on change of mood prediction. The change of mood was predicted based on the relationship between a mood, situations and facial expressions. Situations of the games and recognized facial expressions of the subjects are expected to be keys to estimate a mood based on the relationship.

In recent years, the number of care workers are not enough by a rapid low birth rate and longevity. Therefore, people of the caring at home increase in Japan. The situation are expected to be serious in the future. The robot of this study can alleviate such problems. The social robot is characterized by being able to use all the functions in the operation of the conversation. The information obtained in conversation with this robot pass to any third person through the Android application. In this way, it is possible to easily manage the user information. The conversation features and Android applications are developed in this study.

When they feel familiar to a product, users repeat touching to it frequently. This can be seen in various scene of life. In order to make users feel more familiar to a product, decorating various elements according to users' intention and adding interactive function reacting with users' action are eãective ways. We attempt to develop a communication device to watch their lives by logging their touching behavior to its device. In order to make the communication device have higher familiarity, We developed a module that can be decorated by user as well as interactive function for the communication device. In this paper, we described the module design and its interaction function embedded in the module. Then these function was evaluated via a workshop.

Keeping a living and working spaces tidy is very important for healthy daily life. The efficiency in working rises when the office is tidy. We developed a system that tidies up through the cooperation between a robot and a human. In this study, we investigated the influence of a robot's behavior on the motivation of tidying up. The robot gathered the scattered items on the table to an area in the experiment. We conducted a preliminary experiment with 6 male and 2 female participants.

This paper presents a health promotion system for people with limited health literacy using robot partners. Nowadays, in order to avoid any serious illnesses or overweight health promotion has become an extremely crucial task. For the sake of rising health care intervention, health education should be provided to improve people`s motivation to exercise. In the proposed health promotion system, the content of quizzes provided by robot partner is produced based on a theory of the transtheoretical model. The quiz categories are selected by a stochastic model and they are sequentially updated through the human-robot interaction. In this paper, we introduce an experimental result of a questionnaire to discuss the effectiveness of the proposed system.

Recently a communication robot is being focused on more than ever. While people are communicating with a robot, people are impressed by the appearance of the robot as well as the conversation itself. The appearance includes size, color, shape and looks of the robot. In this research, height is focused as one of the elements of appearance of a robot. We investigated that the impression changed according to the height of a robot, while a user is having conversation with the robot. An original robot, Multi-legged Giant Yardwalker(MGY) is used for the experiments. MGY is able to change its height. The terms such as ’affinity’, ’presence’, ’conversableness’, and ’listenability’ are investigated. Three levels of height, 10%, 45%, and 70% of users’ height were used in the investigation. As a result, 70% got the best result for every item.

Many interactive robots have been developed by many researchers or companies. They can do communication with people and some of them use their hands for expression of their feelings. However, there are few interactive robots which can work. The speech recognition system and the hands of the social robots are explained in this study. The speech recognition system can recognize voice using Google Speech API in any timing. The finger part of the robot hand is produced for picking objects.

Androids with human-like appearance are widely used as communication robots. Android motion is preferred to be human-like. To generate body motions for an android, human motions are directly measured with a motion capture device and transformed into motions for an android. In this paper, we propose a method to generate upper body (neck, torso and arms) motions for an android using a Kinect v2, which is used to measure human motions without markers.

ICT education at schools is to make use of the information and communication technology such as electronic boards or tablets in class through Japan recently. The two different kinds of communication robots have been developed so that people could learn how to lead a better communication using robot technology. Mug-Bot is the social robot on the platform for open sources. Date-Bot has been developed to let the historical personality make dramas by using voice function in social studies class. Connect robot has been developed for the students who cannot go to school due to illness or school refusal to see the movie and the texts of each subject using video monitoring function. In addition, we surveyed, giving questionnaire to Japanese, Chinese, and Korean students and examined the utilization of the robot. In this paper, we describe how to develop, introduce and apply the communication robots in the field of school education.

In this paper, we propose an optimal viewpoint selection system for monitoring robots to search for the optimal viewpoint of a scene with the highest aesthetic property. We first put forward a novel aesthetic evaluation method by the use of Kullback-Leilber divergence, considering the directional information of each target according to some famous composition rules in the field of photography. Then based on the evalution results, we propose a viewpoint selection method by Kullback-Leilber Divergence Minimization. At last, the effectiveness of our optimal viewpoint selection system is confirmed with experiments.

In recent years, the number of elderly people living alone has a tendency to increase due to the advancement of an aging society. Therefore, monitoring and support systems are necessary. In this paper, we developed the therapy robot having monitoring and support function under the consideration to make it not to be a surveillance system. And we evaluated whether it is a suitable size and weight for elderly people. As a result, the answer that size and weight are just right accounted for the majority.

In recent years, coral of significant value is damaged. We have been researching and developing an underwater robot for coral reef conservation. Underwater robot will clean the crown-of-thorns starfish in the acetic acid injection. In this paper, we describe the development of hovering control of underwater robot. Underwater robot uses a hovering on image processing to operation assistance. Image processing use a particle filter. Underwater thrusters will feedback control of the calculation result. Lateral movement control is a control for moving to just beside. Hovering using the image processing will experiment in water tank. Water tank experiment uses a model. Underwater robot was injected by hovering using the image processing in a water tank. Lateral movement control will experiment in pool. Underwater robot moved by lateral control just beside.

This paper describes automatic tracking system to underwater object position measurement. We have been developing the maritime mobile robot for underwater objects position measurement. Position data is one of important information for researchers to research the sea. But it is not easy to tracking underwater objects to operation by person. The authors are going to illustrate developing position measurement system for underwater object using tracking system. The authors show a behavior experiment and tracking experiment.

In this work, we propose a new localization method for underwater robots that can survey narrow environments. The proposed method is based on 3-D map matching, by referring to environmental maps generated from design data and structural shapes measured by an ultrasonic sensor. This paper describes details of the method and calculation algorithms for using an ultrasonic sensor as a structural shape measurement sensor. Additionally, we present experimental results to evaluate the performance. To realize the proposed method, the signals reflected by the structures that include multi-pass signals are used for the environmental information measurement using an ultrasonic sensor. For environmental map generation with multi-pass signals of an ultrasonic sensor, the 3-D trajectory of the ultrasonic wave is calculated by applying the ray tracing method, an approximate calculation method in which the ultrasonic beam is approximated by bundle of acoustic rays. We confirm experimentally that the localization error was decreased in case of using an environmental map considered multi-pass signals of the ultrasonic sensor.

We have developed a drifting float having a buoyancy adjustment mechanism that can maintain the target depth. This float is measured by the GPS position at the time of surfacing. Although GPS is not available in water, it is possible to estimate the flow rate of the target depth from the position of before and after diving. It shows the experimental results in the vicinity of the Kuma-gawa River estuary of Yatsushiro Sea.

In this study, we are developing an underwater robot for surveying radioactive materials in lakes. However, the surveying range is limited by the large weight of robot as well as the capacity of the battery. To expand the range of surveying, the application of underwater glider to the robot is assumed to be effective. This paper describes the basic characteristic examination of the wing through experimentation and simulation as well as the establishment of a position measurement method for an object moving under water.

The autonomous Underwater Vehicles (AUVs) are developed to contribute various science filed such as oceans biology, geology by deep-sea observation. The deep-sea observation by AUV provides various information such as lifestyle of deep-sea creature, volcano of deep-sea bottom, etc. by photographing image. Next mission using AUVs is required to sample targets such as creature, mud or stone in deep-sea floor. For success of this mission, AUV needs supports by researcher in support ship. However, underwater communication based on acoustic has less density compared to electronic. In this research, we proposed a method of image compression and reconstruction for acoustic communication. And we evaluated reconstruction image by Structural similarity.

In this paper, Manta-like robot is proposed to perform the fin with flapping and feathering motion. In the case of rigid fin controlled by feathering angle, stopping performance is evaluated by time response from the initial swimming velocity for flapping frequency. The swimming velocity increases for the frequency of the flapping motion. After reaching the maximum velocity, the feathering angle is controlled in the decreasing velocity period. The result shows that the feathering angle control for the Manta-like robot is effective to stopping motion in the water.

Master slave system is a vital technique for controlling robot motions. An intuitive master controller enables precise and easy control of robot, especially for underwater robotics applications. We are developing a master slave system for a 3-link dual-arm underwater robot. The design of this master slave system is simple and intuitive. Since to correctly operate the underwater robot is difficult, it needs operator training. In this paper, we discuss a master-slave simulator for 3-link dual-arm underwater robot.

Underwater robots are employed to marine organisms' research in ultimate environments. In this study, we aim to do automation of marine organisms sampling by autonomous underwater vehicle with a manipulator. We developed a manipulator to pick up deep-sea benthic organisms. The developed manipulator is driven by a master-slave cylinder type hydraulic driven system. Moreover, the manipulator's joints are actuated by hydraulic cylinder, and linear motion of the cylinder is converted to rotational motion by a motion conversion cam-mechanism. This control system was integrated by MATLAB/Simulink. So we adopted Stateflow (this is a toolbox for Simulink) in the high level control layer.

Currently, many ships and underwater vehicles use ballasts to control their buoyancy, although it is harmful to the environment. We are developing a volume variable buoyancy control device using the volume change induced by phase transition of material. In our previous study, we developed two-step piston mechanism to increase volume change. However, the mechanism had a drawback that the piston could not return to the initial volume after increasing the volume. We, then, come up with a new idea of volume change mechanism which combines a two-step piston and a metal bellows. We manufactured a device and conducted some experiments on its specification. As a result, we confirmed that the device can change its buoyancy.

For long-term observation of the seafloor by AUVs, it is required to charge its battery. To charge battery of AUVs, we had developed autonomous docking method by using the seafloor charging station. This paper proposes the robust docking method under the environments with limited visibility or the sea current. The method consists of two techniques. One is the state estimation using a particle filter. The AUV recognizes each LED and estimates its state based on measurements of LEDs. The second is the control technique against the sea current. The method was implemented to AUV Tri-TON2 (TT2). Through the simulation based on the data obtained from sea environments, and tank tests using TT2, the performance of the method was verified.

Underwater is a challenging work environment with high risks for human beings. One of the needs of the robot, the development of robots working in water extreme environment is expected. Autonomous Underwater Vehicles (AUVs) are suitable for a wide range of search mission without restrictions from power cable, AUVs can move freely in every direction. In this research, we have developed a SSBL acoustic positioning system for underwater robot. Also we report the experimental results of acoustic positioning.

Master-slave system is a vital technique for controlling motions. We confirmed usefulness of master controller for an experimental semi-autonomous underwater vehicle equipped with 3-link dual-arm. To execute tasks such as handling and catching object underwater, we produced manipulator hands for using underwater. In this paper , we will describe the result of object capturing experiment of the underwater robot using master-slave system for UVMS.

This paper proposes a method for long-term continuous deployment of a hovering type AUV based on autonomous docking to a seafloor charging station. The method consists of precise navigation method based on both acoustic and visual landmarks attached on the station. The AUV stochastically estimates its position and orientation with regard to the station and controls itself to land on the exact docking position of the station. The AUV Tri-TON 2 succeeded in autonomous docking to the testbed station in both a test tank and real sea. We had also charged the AUV during docking by our newly developed non-contact charging devices.

Recently, ocean energy conversion systems using offshore structures such as mega-floats are been developed. Stable operations performed of those structures require constant monitoring and maintenance. In this study, we propose position and posture control methods using acceleration sensors and depth sensors for an Autonomous Underwater Vehicle (AUV) for maintenance of underwater structures. We made a small underwater mobile robot with sensors and thrusters for 3-D motion in the water. We performed experiments for considering performances of position and position controls.

This paper describes that the pilot experiment of the multicopter contacting vertical and horizontal surfaces. Our multicopter with one DoF manipulator has been being developed to enable to execute an inspection of aged bridges. For the inspection, the multicopter shold be able to execute hammering check and visual check. Our current study focus on the hammering check. To execute the hammering check stably, the multicopter should exert contact force against target surfaces. In an experiment, we measured the contact force when the multicopter contacted vertical and horizontal surfaces. It was confirmed that the measured contact force was proportional to the pitch angle of the multicopter and the throttle rate in each contact condition. The tendency of measured contact force followed numerical values. Therefore, it was confirmed that referential attitude and throttle rate for contact-force modulation could be calculated from this relationship.

The deteriorations of the social infrastructures such as the bridges become a serious problem. Therefore, the inspections of the infrastructures are necessary. Because of the dangerousness of these inspections carried out in high place, inspection robots based on multirotor with image inspection and hammering test are widely developed. However, it is difficult to carry out the two tests in a multirotor. Thus, the multirotors with adsorption mechanism to wall surface have been proposed based on two ideas; vacuum sucker or claw. However, multirotor using vacuum sucker can adsorb only a flat wall, and multirotor using claw can adsorb only wall having a step. In this research, we focused on one of the vacuum sucker, called Universal Vacuum Gripper, in short UVG. UVG is vacuum sucker adsorb to uneven surface with deformable skirt. In this paper, we propose three multirotors based on UVG which adsorb vertical wall and ceiling. Here, we report the performance of the wall surface and ceiling adsorption by the multirotors.

This paper presents the method for enabling an Unmanned Aerial Vehicle (UAV) to attach on walls and ceilings. Recently UAVs are expected to utilize for gathering information and work at high altitude. However, continuous flying consumes a lot of energy. Therefore attaching to wall and ceiling is effective function for saving energy. We developed a hybrid suction cup combined with a sucking disc and a sheet of gel. It can generate both the adhesion and the absorption, which allows UAV to attach on wall, plaster board, tile and so on. However, it is also necessary to tear off gel in order to leave from the attached surface. So we added an actuator named Release Balloon for desorption. Then we verify the effectiveness of our method through experiment.

A quad tilt rotor UAV has ability to fly with completely independent position and attitude control by tilting the directions of the propellers. However, the control method against disturbance with this type of UAV has not been considered yet. This paper proposes disturbance detection method using IMU sensor and develops attitude recovery flight control method against impact disturbance for the quad tilt rotor UAV. The numerical simulation and fundamental flight experiments were carried out to evaluate the efficiency of the proposed attitude recovery flight control.

This paper presents the realization of sustained flight of a flapping wing robot. Unmanned aerial vehicles (UAVs) have attracted a significant amount of attention in recent years due to their ability of flight. One of the UAVs attracting attention is the flapping robot like a bird. The robot can fly in the sky, and is safer than rotorcrafts such as helicopter and multicopter since the operating frequency is much less than rotorcrafts. In our previous paper, flapping wing robot with oscillating controlled RC servomotors has been developed and succeeded in a flight of 30m. However, long-distance and sustained flight is not realized yet. The purpose of this research is the realization of sustained flight of the flapping wing robot. We redesign and improve the flapping robot. Moreover, we make a catapult for achieving reproducible flight experiments.

This paper describe a tricoper using the duct fan. A tricopter is kind of Unmanned Aerial Vehicle(UAV), having three propellers. Its yaw angle is controlled by using tale tilting mechanism. It is downsized and high thrusted by replacing propellers of tricopter with duct fun. Through a hovering experiment, the effectiveness of the mechanism was verified.

Small unmanned aerial vehicles (SUAV) have recently attracted considerable research and development interest owing to their small size, portability, and limited operational risk. In SUAV, the Reynolds number is typically very low because they are small and fly at low speeds. The Re strongly influences the airfoil performance, and at Re = 10⁵, the lift-drag (L/D) ratio decreases significantly and the airfoil performance deteriorates. Since SUAV must be able to carry payloads such as communication devices and observation equipment, the challenge is to maintain sufficient payload capacity. Therefore we proposed the Circulation-controlled high-lift wing that can maintain a high L/D ratio and increase more payload in the low Reynolds number region (Re = ～105). In this paper, we conducted with the flow visualization test to verify the prevent effect of the peeling of airflow by the smoke-wire method.