The study has proposed a multi joint gripper (MJG) that achieves envelope grasping for unknown shape objects. The mechanism mainly consists of a serially connected differential gear system (DGS) controlled by only two actuators. We have developed the MJG with seven joints and succeeded the envelope grasping of various shape objects with no sensory feedback. This paper proposes the way to control the joint stiffness followed by being elucidated with some simulation studies and experiments.
To construct an intelligent space which has a distributed camera network, pre-calibration of all cameras (i.e., determining the absolute poses of each camera) is an essential task that is extremely tedious. This paper deals with automatic calibration method for the distributed camera based on 3D map information of an environment. Parameterized line features that are extracted from both a distributed camera image and the map information are transformed to Hough space and utilized for matching process in particle filter-based estimation. We evaluate the proposed method in a simulation environment with a virtual camera.
This paper describes a simplified sensing system for estimation of dynamics of ride-type interface such as the driving simulator and tele-operated construction machine for motion generation. The dynamics of the interface of body of equipment and the condition of drive housing of the operators are needed for motion generation of the ride-type interface. We construct a simplified measurement system of the behavior of the drive housing of ride-type interface using a micro controller. Experimental results show the feasibility of the simplified measurement of the behavior of ride-type interface.
We demonstrate the effectiveness of space-time features incorporated in person re-identification (Re-ID). The space-time representation, improved dense trajectories, effectively extracts multiple features on a large number of trajectories. The experiments show that our proposal outperforms convolution neural network features on the iLIDS-VID dataset. Moreover, we customize the space-time feature for Re-ID. The numbers of temporal accumulations and visual words are highly tuned to classify individuals.
In this pilot study we are investigating the feasibility of using the humanoid robot Pepper to monitor and encourage the elderly to walk, making the interaction safe and enjoyable at the same time. The results of this first experiment will lead to improve the design of an accompanying robot that could help the overworked staff in a nursing home scenario.
This paper presents an Inertial Measurement Unit (IMU) sensor analysis system for manipulator-based position tracking system. Later, the position tracking system will be used for a three-dimensional ultrasound (3DUS) imaging application. To obtain better accuracy, the proposed experiment was done using 3D-printed gimbal mechanism. Comparison between accelerometer, gyroscope, complimentary and Kalman filter reading was done using 3 different rotations; roll, pitch and yaw. Results show that complimentary and Kalman filter are better than the accelerometer and gyroscope reading.
In recent years, demands for outdoor mobile robots, such as those for agriculture or environmental monitoring, are increasing. Power supply is one of the most important considerations for these robots, and autonomous battery charging system can be one solution. The objective of this study is to develop a prototype battery charging system for outdoor mobile robots, using wireless power transmission technology. We developed a battery charging station and a power receiver for a mobile robot. The receiver coil was implemented in WAMOT-2, which had been developed for the monitoring. Operation of the system was verified through an experiment.
This paper describes an accurate attitude determination technique using multiple global navigation satellite system (GNSS) receivers and antennas for small unmanned aerial vehicles (UAVs). A laser survey using a small UAV needs highly accurate attitude information for reconstructing 3D models. The proposed technique uses the multiple GNSS antenna's positions computed by the GNSS carrier phase measurements to estimate the UAV's attitude. To solve the GNSS carrier phase ambiguity, the reliable ambiguity search is performed with a couple of constraints for the base line vector. The performance of the proposed technique is evaluated in the static test. From the test, the proposed technique could estimate UAV's yaw angle in 0.1 degree accuracy.
In recent years there has been interest in the development of magnetorheological actuators for robotic applications with adjustable compliance. In this paper we explore the conceptualization, design and test of a new kind magnetorheological piston head design. Inside the head of this new damper a circular assembly of magnetorheological valves is used to control the damping force of the piston. The results show the feasibility of the new design, and its potential for compliant robotic applications.
We propose a conceptual framework of smart mechanics with novel personal mobility systems. The developed device is able to assist user's voluntary postural change: sitting to standing with a passive mechanism, which also allows to support the convenient transferring from the wheelchair into a bed, toilet, tub or a regular chair. In addition, we proposed a step-climbing wheelchair also with a passive mechanism without using any electrical motors. Potential target users of both systems are youth with spinal cord injury (SCI) who face daily problems in typical living environments, which do not always have ramps or elevators, corridors are relatively narrow, shelves are high, often upright eye level or even higher.
Teleoperated construction machine begins to be introduced in industrial uses. When the operator controls construction machine, the lack of depth perception makes the working efficiency worse. We focused on a depth cue of motion parallax that is effective for operating construction machine. The motion parallax display system is developed to give the remote operator depth sensations.
Recently, improvement in robots with human affinity is necessary. This improvement can be achieved by designing a telescopic actuator comparable to that of a human muscle cell. We study a shape memory alloy (SMA) actuator protected by a rolled film tube. In the present study, the actuator is inserted into a natural rubber tube and investigated the characteristics.
This article reports the current status of standardization of wearable robots in ISO and IEC. ISO focuses on the non-medical purpose; on the other hand, IEC focuses on the medical purpose due to historical and political reasons. This article will help to outline the current direction of the standardization.
Bipedal Walking Robot uses harmonic gears in its leg joints in order to provide high torque. However, those mechanisms increase the leg's weight that automatically increases the required energy. Replacing the drivable gear system with an adjustable stiffness mechanism might help to increase the energy efficiency. In this research we purpose a novel adjustable stiffness mechanism that can be attached to the ankle joint. The developed mechanism can provide passive and active controlled motion, which might help the robot to walk with less energy.
Robotic technology in the form of master-slave system provides dexterous motions in the confined workspace while increasing the accuracy in motion. Surgical robot is one example that the robotic technology has been successfully introduced in the market, and its applicable area is currently increasing. As an extension of the concept, we develop a laparoscopic manipulator to perform an ultrasonic scan in the abdominal cavity. The prototype was built as a hand-held device with three degree-of-freedom that can perform a dexterous ultrasonic scan in the confined abdominal cavity that has been difficult using conventional surgical tools. In this paper, we introduce the mechanical structure and its evaluation.
The aim of this article is to describe the design and locomotion of eight-legged soft mobile robot. Most of soft robots are implanting solid-state structure as an actuator for locomotion and/ or manipulation. Although few still harnessing conventional actuator from pneumatic and hydraulic, the response is sluggish. Here, we design exclusively soft body mobile robot without solid-state structure with fast and flexible mobility of locomotion. The robot is made from silicone rubber and operated pneumatically. Experimental results show the ability to perform flexible locomotion with locomotion speed of 5 mm/s to 6.9 mm/s at different locomotion direction. The features offer some prospect for the robot to be used in medical and industrial automation.
In this paper we use an extended footstep planning algorithm to plan optimal humanoid locomotion trajectories subject to constraints on the maximum predicted Zero Moment Point (ZMP) tracking error. The approach can guarantee walking stability bounds with little extra computational burden, thus increasing safety of robots walking in challenging environments. This is done by estimating energy and stability models in simulation through Bayesian optimization, and smartly integrating the models into search-based planning.
The paper describes a method of extraction of grasping region of objects for automatic modeling by observing human-object interaction. Human grasps and manipulates objects in everyday life. Home service robots and environmental cameras in intelligent space may obtain object models by observing such human actions. The proposed method extracts grasping region of object by observing human-object interaction such as grasping and manipulating objects from the first person point of view with head mount camera.
We propose a simple method to estimate fish-eye camera's intrinsic parameters without calibration targets. Our method takes advantage of trajectories of feature points in the scene. The trajectories of feature points are obtained by a rotation movement of the camera in a horizontal plane. We therefore can utilize rich feature points for calibration, and furthermore, specific calibration targets are not required. The validity of proposed method is verified by simulation experiments using artificial data.
This article describes about mobile sensor nodes (MSNs) deployment method to maintain communication connectivity for constructing a disaster information gathering system called Robot Wireless Sensor Networks (RWSN) we have been discussing, and also describes the result of communication quality evaluation by using MSNs in an actual outdoor environment.
Giacometti robotics have the potential to solve problems related to safety for practical uses in a way that is very different from conventional robots. Based on this new concept, our research group is developing a hexapod Giacometti robot with very long, light, and thin legs. In this paper, we introduce a concept of hexapod Giacometti robot and report the designe of leg mechanism, and its basic performance by using the prototype. Prototype successfully demonstrates basic walking motion.
In this study, we develop a prototype telepresence childcare robot, which will aid childcare by facilitating interactions between a young child and a remote person. Because the robot interacts with a young child, safety measures and an ability to maintain the attention of a child are important considerations. We design a robot with a focus on safety and physical play, and propose the use of a semi-autonomous child tracking method. Finally, we perform a preliminary test, which indicates that this robot can maintain a child's attention for longer than a general video-chat system.
This paper studies retrieving geometrical features of objects with hollows and motion planning of caging constraint for such objects. To achieve caging a ring-like object, position of hollow parts is essential information, which can be acquired by vision system. With basic image processing methods, we can detect the features and posture of scissors, which has hollows on its handle. Motion planning for caging a pair of scissors can be successfully achieved within reasonable computing time.
This paper describes the development of a four-limbed robot capable of climbing vertical ladders. We focused on vertical ladders which are often set in plant facilities. The design concepts and required specifications are explained. We conducted experiments of climbing vertical ladders and the robot succeeded in climbing vertical ladders in JIS.
This article discusses functions of a stationary underwater sensor node (S-USN) with attached algae removal ability for constructing underwater monitoring sensor network (UMSN) that is used to support coral ecological system survey activities. The experiments of attached algae growth was conducted in Okinawa coast. Based on knowledge obtained by the experiments and coral researcher's advice, container surface wiping function was implemented to S-USN.
In this study, we developed a pneumatically-driven manipulator for the master-slave-type surgical robot. The manipulator has dual flexible joints and be capable of bending the tip in two degrees of freedom by two pneumatic cylinders. Considering the interoperability in real surgery, we proposed a mechanism so that the sterilizable forceps part can be easily attached to and detached from the nonsterilizable drive unit. To verify the tracking performance, we actually conducted an experiment of the master-slave-system with the proposed manipulator.
In this paper, a device to control a pre-stress between a rotor and a stator of cryogenic ultrasonic motors has been introduced. The friction force between the rotor and the stator is an important factor to drive the ultrasonic motor. We have fabricated and evaluated the device to control the friction force by using piezoelectric actuator in cryogenic environment.
The accuracy of side-slip angle and yaw rate measurements is important in vehicle motion control. However, vehicle side-slip angle measurement is difficult. In general, side-slip angle can be estimated by observer. In this paper, we propose a vehicle state estimation method based on Moving Horizon Estimation (MHE) using acceleration sensor and gyro sensor. Measurement data of inertial sensor has large noise, since S/N ratio is small, especially if estimation object is a light weight vehicle. Moreover, outlier of yaw rate measurement value is caused by road surface disturbance. MHE is used for vehicle state estimation that can suppress the measurement noise on inertial sensors. In this paper, we show the efficacy of the proposed method through the simulation.
We present two sets of outdoor LiDAR dataset for semantic place labeling using two different LiDAR sensors. Recognizing outdoor places according to semantic categories is useful for a mobile service robot, which works adaptively according to the surrounding conditions. However, place recognition is not straight forward due to the wide variety of environments and sensor performance limitations. In this paper, we present two sets of outdoor LiDAR dataset captured by two different LiDAR sensors, SICK and FARO LiDAR sensors. The LiDAR datasets consist of four different semantic places including forest, residential area, parking lot and urban area categories. The datasets are useful for benchmarking vision-based semantic place labeling in outdoor environments.
Underwater multiple observation system has developed it for the purpose of doing seabed resource-searching efficiently. In this paper, a process of the development of an underwater multiple observation system, equipment configuration and an operation procedure are described.
This paper describes a head position estimation method for a variety of recumbent positions for a care robot. First, the robot detects a fallen person using a laser range finder and moves there for getting a point cloud. The obtained point cloud is matched with those with head positions in the database using the ICP (iterative closest point) algorithm for estimating the current head position. Then, extracted head positions with matching scores above a threshold are clustered, and the center of the cluster with the highest accumulated scores is determined to be the head position. A human pose database construction system is also described.
This paper proposes and discusses the effects of robot edutainment class for bachelor students by using robot kit. The aim of the class is to improve their understanding and ability about robot development. This class enhances their skill by exerting "imagination and creativity" and "thinking and attempting" through robot development with other students and robot contests. We have been practising the class for four years. For several years, we have changed curriculum of the class to enhance the quality of the education. We discuss and evaluate the effects of the class by comparing the developed robots by the students with changing the curriculum.
Locomotive performance is one of the most important issues for land-based mobile robots, which is designed to be used in natural environment, such as grass fields or forests. We propose a novel method to change outer shape of the robot to increase locomotive performance in natural environment. The robots, which are designed using this method, does not require additional actuators, while conventional mobile robots obtain higher locomotive performance in natural environment using multiple actuators. The objective of this study is to propose a novel design method of outer shape of mobile robots, which is inspired from "icebreakers". Through experiments, we confirmed that locomotive performance is greatly affected by outer shape of mobile robots.
This study aims to investigate the relationship between elderly monitoring system data and climate condition, and to investigate the difference of reaction to the climate condition caused by the cognitive function, motor function and self-rated health. Four pyroelectric sensors are set at 15 elderly persons' home, and the sensors can detect the subject's movement in their house. Correlation between their total activities at home and the climate conditions is calculated considering the difference by the health conditions. Correlation between total activities and average temperature showed moderate correlation (R=-0.41). Comparing between higher cognitive function group and lower cognitive function group, high group showed moderate correlation (R=-0.43), but low group showed little correlation (R=-0.25). On the other hand, in high group, their total activities are not much changed by weather, but in low group, their total activities in rainy days are about 5 % lower than that in sunny or cloudy days (p=0.05). These results give us suggestions about seasonal changes or person specific patterns are related to the change of the cognitive functions. More information about the activities in or out of their home may help earlier detecting the change of health conditions.
We developed a mobile power assist robot to be a human-coexistence robot. The present paper discusses the development of a power assist arm and the moving mechanism of a power assist robot. We report the mockup of a power assist robot.
In this paper, a development of a disaster response manipulator having a variable stiffness joint in shoulder will be proposed. The manipulator is aiming to realize lightweight and transportable reconfigurable robots that can be operated by human workers at disaster sites just after the occasion of events. To demonstrate the feasibility of the proposed design and conceptual tactile sensing algorithm, a motion control experiment to measure the movability of a specimen will be performed by using a fabricated prototype.
In this paper, we present an investigation of motor imagery (MI) based brain-computer interface (BCI) performance with visual and vibrotactile feedback in participants with different MI abilities. We tested the MI ability of 18 young adults and performed training sessions with 4 of them. The outcome showed a positive effect of the addition for vibrotactile feedback, but none for MI ability, contrary to our initial hypothesis.
Autonomous mobile robots has been an active research recently. Especially, laser range finder (LRF) based navigation has rapidly improved. On the other hand, in spite of intensive studies on vision based navigation using cameras, they are not successful in real world. In this paper, we develop a vision based navigation method using topological approach to reduce prior information and enhance scalability of the map. Topological map has advantage in scalability because it contains only information about impressive places (i.e. intersection) and their linkage with roads. The robot is navigated using two primal techniques: vision based road following and impressive place recognition.
Existing teleoperated construction machineries are by far less efficient than directly operated ones and principal factors of inefficiency of existing systems have not yet clearly determined. The authors are developing a teleoperability evaluation platform that is a prototype of a next generation user interface for a remote operator. The platform can provide rich surrounding information of an teleoperated subject to extract the principal factor that degrade work efficiency. Moreover, some next generation systems that improve work efficiency of teleoperation are tested on the platform. This paper describes an outline of the platform and the systems.
Automation of the scooping of rock sediments by wheel loaders in quarries requires the model of reaction forces from rock sediments consisting of various sized rocks. In this paper the reaction forces are formulated based on Coulomb's earth pressure model and it is verified by fundamental experiments with a mechanical simulator and a video camera.
In this study, a gamma irradiation experiment for our proposed bird's-eye view system was conducted. Experimental results show that cameras attached to our system malfunctioned one by one in the irradiation environment, and the proposed system generated bird's-eye view images until all cameras malfunctioned.
Using ultrasonic motors, we propose a compact and highly responsible eye robot that is moved by transmitting a operator eye movement to the robot's eye, to obtain the visual information of the environment surrounding the robot. In addition, we propose a graphical user interface (GUI) that makes the robot operation intuitive.
In this study, we developed a manipulator that induces a horizontal surface rotation motion of the shoulder to change the phase difference between the shoulder and the upper arm. Furthermore, we constructed an algorithm that estimates the time of heel contact during walking on the treadmill and the angle of the shoulder in the axial rotation was controlled. Accuracy of the proposed algorithm was verified through experimental data. As a result, the angle could be controlled with as much accuracy as in young healthy individuals where the angle varies for each gait cycle.
We have been developing a lunar subsurface excavation robot using peristaltic crawling based on an earthworm's locomotion. The robot is designed to discharge the excavated soil from the borehole to the ground surface. However, the soil-discharging process is not yet automated. This paper develops and tests two promising soil-discharging mechanisms for automatic soil discharge.
A method for a six-legged robot of walking on deeply undulating terrain by tripod gait is proposed. "Deeply undulating terrain" is defined as ground with large difference of elevation; its shape is unknown. The robot inclines its body to make its legs put on the ground. First the robot detects foot landing using touch sensors, and estimates the inclination angle of the ground from the landing positions; the estimated ground is called "reference ground". Second the robot measures the inclination angle of its body using acceleration sensor. Then the legs move up and down so that the body may be parallel to and may keep a constant distance from the reference ground. If the robot encounters deep depressions or downward slopes, the reference ground is actively inclined. As a result the body inclines, and the legs can reach the ground. The ability of climbing up and down slopes and steps is evaluated by experiments.
Towards the actualization of an autonomous guitar teaching-aid system, this paper proposes the following two video analysis based methods: (1) guitar neck detecting and (2) guitar neck tracking. For (1), an algorithm that can extract the rectangles cut by guitar strings and fret is proposed so that we can detect guitar neck area on the first frame of the input video. For (2), a feature point based guitar neck tracking algorithm is proposed, which can cope with occlusion issue effectively by analyzing geometric relationship between each two feature points in every frame. Experimental results demonstrate that the proposed two modules are robust enough under complex contexts such as complicated background and different illumination conditions.
We developed a flexible sheet-type tactile display 'Haptoskin' with a CNT electrode that can be attached to objects and decorate the objects with extra tactile characteristics, such as smooth and rough. When the voltage is applied to the electrode, the finger on the PDMS layer is attracted to the electrode and experiences vibration-like tactile sensation. We fabricated Haptoskin with MEMS-process friendly materials, PDMS and CNTs. We attached the Haptoskin onto a flat and curved surface and requested subjects to answer the minimum voltage when they could recognize the tactile sensation. In both cases, the Haptoskin successfully presented tactile sensations to the subjects. The developed Haptoskin can be readily used as a new human interface device.
Electrical muscle stimulation (EMS) has been utilized in various haptic interfaces. To enhance pseudo-haptic feedback provided via the EMS, we have proposed to incorporated visual feedback with the electric stimulation and confirmed the effect of the feedback on the sense of force induced by the EMS. In this study, we investigated the influence of the vibrotactile feedback in addition to the visual feedback on the haptic sensation by the EMS. The experimental results show that the perceived force with the visual and the vibrotactile feedback tend to be larger than the force exerted in the sole electrical stimulation condition.