This paper presents a new method of realizing high-speed bipedal running with a high-speed robot using high-speed visual feedback. First, to achieve high-speed running with simple control, we developed a biped robot composed of lightweight high-torque actuators. Next we propose a control strategy which does not depend on typical methods based on a zero moment point, but instead modifies the robot's posture by high-speed visual feedback. In the strategy, by adding instantaneous high torque, we confirmed that the robot could execute running in the unbalanced region. We realized bipedal running at the speed corresponding to a Froude number of approximately 1.0, the highest reported value for non-birdlike bipedal robots.
The portable air supply system for wearable device had been developed in previous works. The previous system, which can retrieve compressed air from an actuator, is effective to decrease energy consumption. However, supplying / retrieving compressed air into / from an actuator and generating compressed air by compressor can not be performed simultaneously because energy flowed into / out a tank can not be measured directly. Establishment of estimation of pneumatic energy is desirable in order to control inner pressure of actuator continuously in future works. Air flow must be required to calculate pneumatic energy. However, the system becomes large by introducing air flow sensors. Air flow passed through an orifice plate can be calculated by both sides pressure of an orifice plate. Accordingly, air flow can be calculated by assuming an air valve to be an orifice plate. Real-time energy estimation will enable to supply / retrieve compressed air and to generate compressed air simultaneously. In this paper, the energy estimation method is discussed, and then the operation flow for this system based on pneumatic energy is described. Finally, experiment using actual system is performed to verify the proposed method.
This paper develops a wearable walk rehabilitation device for hemiplegic legs with facilitative vibration stimulus and power assistance, based on Repetitive Facilitation Exercise, an effective rehabilitation method that can potentially manifest the brain plasticity, which can realize rehabilitation that incorporates facilitated vibration stimulus and active power assist for hemiplegic legs. In this research, hip and knee joints power assist mechanism, 3-DOF hip joint and 1-DOF knee joint rotation adaptation mechanisms to the joint rotation of the human body, mechanism corresponding to the individual differences in body shape, and mechanism to prevent side-overturning preventing mechanism are proposed. The designed prototype of the device, then the effectiveness of this device are shown from some verification experiments.
This paper proposes a new method for knotting a rope with a dual-arm robot system. A knotting task is divided into a series of “steps” taking account for a timing of getting visual information. The start state of each step will be observed and confirmed visually. And an adequate operation from the start state to the end state (i.e. the start state of the next step) is conducted according to the visual information. The method was implemented on a dual-arm robot. Five different types of knotting tasks were realized with the robot system. These knotting tasks have common steps each other and the steps were successfully reused. The experimental result showed that complicated knotting tasks could be realized as a sequence of these kind of steps.
This paper proposes a method for imitating human regrasping motion by a robot. The method is based on the learning from observation (LFO) paradigm, in which human motion is recognized by a task model and robot motion is reproduced from the recognized task sequence. For designing a task model for regrasping, we focused on a topological criterion, Gauss linking integral (GLI), which represents a tangle state of two strands. Fingers and an object are represented by strands in this study, and the relationship among them is described in a topological way based on GLI. In this paper, first, a task model for regrasping is proposed, and then, a method for recognizing human regrasping motion using the task model is described. Next, the proposed method is validated by reproducing a regrasping movement in a robotic hand. Human regrasping movements of a pen-like object are considered. The successful reproduction of the regrasping movement verifies the proposed methodology to be useful and proved that it is feasible to control a robotic hand by imitating human.
Mimetic soft underwater robots using PFC are being developed by author's group. High speed and efficiency motions have been realized by the developed robots up to now. However, the robots are with simple shape of flat plate. To realize stream-line shape and embedding electric components inside the robots, it is necessary to prepare interior space inside the robots. For this purpose, this paper proposed basic structure, and the design approach for 3-dimensional fish-like soft underwater robots. And we confirmed the validity for propulsion performance improvement by experiments using developed prototype designed by the proposed method.
Getting a satisfactory maneuverability in a power-assisted mobile robot is the most important subject. The difficulty in this subject is in a difference of the characteristics of operators. This research focused on the operation of the power-assisted mobile robot with velocity-based impedance control. From the analysis of experimental video, we found that there was a difference in the posture between operators. In order to analyze the difference quantitatively, manipulability ellipsoid which is used for the analysis of manipulators is used. From results, an operator which takes larger size of the manipulability ellipsoid to the moving direction can operate the robot smoothly. Additionally, results show that maneuverability can be improved by adjusting the handle position to become the manipulability ellipsoid larger. However, the structure of a handle cannot be designed freely in almost robots. In this case, maneuverability can be improved by adjusting parameters of the controller depending on the size of manipulability ellipsoid, if it can be measured directly or indirectly.
In this paper, we introduce the bin-picking robot system we developed for the manufacturing industry. The system, that is composed of RGB-D cameras and a dual-arm robot, should locate the target object inside a bin, pick it, and place it onto the assembly jig. As target objects, three types of pipes and two types of flat plates (all of them made of metal) were used. To estimate the position and pose of an object inside the bin, the depth image captured by the RGB-D camera is used. And we describe three evaluation methods; accuracy, reproducibility and availability, those applied for an object's position and pose estimation algorithm, too. The accuracy and the reproducibility of the estimation are evaluated using an XYθ stage. The object is transferred in X and Y directions and rotated about Z axis using the stage, and the estimated transfer matrix of each location is compared to the real one. The availability for the estimation algorithm is measured by simulating the manipulation task with human operation. The results of the estimation agree with the manipulation experiments performed on the robot.
This work presents a human-robot cooperative approach for infrastructure inspection. The goal is to create a robot that assists the human inspector during hammer sounding inspections that detects invisible defects under the surface of concrete by striking the surface with a hammer and listening the resulting sound. The conventional hammer sounding inspection is time-consuming, and there is no convenient way to represent exhaustively the test results. In the proposed approach, an assistant robot accurately estimates the position of the impact in real-time and creates a detailed representation of the test results. Experimental results show the process for creating the detailed inspection report. The accuracy of the human-robot cooperative approach is evaluated for a real world application. The center of the error distribution of the impact point estimation was 44[mm] from the ground-truth with 27[mm] of standard deviation.
Recently, various multi-fingered hands have been developed to perform a variety of tasks like humans. Many of them have a large number of actuators and complex mechanisms, so that they should be expensive and require complex control. The purpose of our research is achieving diversity by simple mechanisms. Human being achieves various tasks with tools, e.g. scissors, nipper, pliers, driver, etc. If robot handles these tools, its diversity might increase. In this paper, we propose a robot hand that can handle these tools by placing two additional passive joints at fingertips of a parallel gripper. In addition, we clarify necessary knowledge to achieve a paper cutting task by scissors with the proposed robot hand. Furthermore, control approaches for paper cutting task are proposed. Feasibilities of our robot hand and proposed approaches are confirmed by experiments.