Eimei Oyama

A complete analytical solution to the inverse kinematics of the Pioneer 2 robotic arm

For robotic manipulators that are redundant or with high degrees of freedom (dof), an analytical solution to the inverse kinematics is very difficult or impossible. Pioneer 2 robotic arm (P2Arm) is a recently developed and widely used 5-dof manipulator. There is no effective solution to its inverse kinematics to date. This paper presents a first complete analytical solution to the inverse kinematics of the P2Arm, which makes it possible to control the arm to any reachable position in an unstructured environment. The strategies developed in this paper could also be useful for solving the inverse kinematics problem of other types of robotic arms.

Inverse kinematics learning by modular architecture neural networks with performance prediction networks

Inverse kinematics computation using an artificial neural network that learns the inverse kinematics of a robot arm has been employed by many researchers. However, the inverse kinematics system of typical robot arms with joint limits is a multivalued and discontinuous function. Since it is difficult for a well-known multilayer neural network to approximate such a function, a correct inverse kinematics model cannot be obtained by using a single neural network. In order to overcome the discontinuity of the inverse kinematics function, we proposed a novel modular neural network system that consists of a number of expert neural networks. Each expert approximates the continuous part of the inverse kinematics function. The proposed system uses the forward kinematics model for selection of experts. When the number of the experts increases, the computation time for calculating the inverse kinematics solution also increases without using the parallel computing system. In order to reduce the computation time, we propose a novel expert selection by using the performance prediction networks which directly calculate the performances of the experts.

Experimental study on remote manipulation using virtual reality

To control a slave robot in poor visibility environments, an experimental extended teleexistence system using virtual reality was constructed. The environment model was constructed from the design data of the real environment. When virtual reality is used for controlling a slave robot, the modeling errors of the environment model must be calibrated. A model-based calibration system using image measurements is proposed for matching the real environment and the virtual environment. The slave robot has an impedance control system for contact tasks and for compensating for the errors that remain after the calibration. After the calibration, an experimental operation in a poor visibility environment was successfully conducted.

Inverse kinematics learning for robotic arms with fewer degrees of freedom by modular neural network systems

Artificial neural networks have been traditionally employed to learn and compute the inverse kinematics of a robotic arm. However, the inverse kinematics model of a typical robotic arm with joint limits is a multi-valued and discontinuous function. Because it is difficult for a multilayer neural network to approximate this type of function, an accurate inverse kinematics model cannot be obtained by using a single neural network. In order to overcome the difficulties of inverse kinematics learning, we propose a novel modular neural network system that consists of a number of expert modules, where each expert approximates a continuous part of the inverse kinematics function. The proposed system selects one appropriate expert whose output minimizes the expected position/orientation error of the end-effector of the arm. The system can learn a precise inverse kinematics model of a robotic arm with equal or more degrees of freedom than that of its end-effector. However, there are robotic arms with fewer degrees of freedom, where the system cannot learn their precise inverse kinematics model. We have adopted a modified Gauss-Newton method for finding the least-squares solution to address this issue. Through the modifications presented in this paper, the improved modular neural network system can obtain a precise inverse kinematics model of a general robotic arm.

Planning Of Landmark Measurement For The Navigation A Mobile Robot

Recognition of its current location is im- portant for a mobile robot to follow a planned path. Landmarks, which are assumed in the environment around the robot, are used to support this recognition. In this paper we propose a method for the robot to select the most suitable landmark among the measur- able landmarks from its current position by evaluating the error covariance update of the position estimation. We use the extended Kalman filter to update the po- sition estimation. We show the update behaviors of two types of landmarks : line type and point type. For the criteria of the evaluation we use the widths of the error ellipses, perpendicular to the direction of the course. Through the simulation of navigating a robot along the center lines of corridors the effectiveness of the proposed method is confirmed. ing landmarks the position estimation is updated in the same way. The behaviors of using two types of landmarks, a line type landmark and a point type one, are examined. The position estimation using a particular landmark is ex- pressed by an error ellipse which is derived from the error covariance update and which shows the directional error distribution of the estimated position error. Though it is favorable if the radius of the error ellipse is small in all di- rections, we think the width of the ellipse along the course is important and propose to use it as the criterion for the selection of the landmark. By the simulation of navigating a robot along the center lines of corridors the effectiveness of navigation based on the proposed landmark selection method is shown.

Modular neural net system for inverse kinematics learning

Inverse kinematics computation using an artificial neural network that learns the inverse kinematics of a robot arm has been employed by many researchers, However, conventional learning methodologies do not pay enough attention the the discontinuity of the inverse kinematics system of typical robot arms with joint limits. The inverse kinematics system of the robot arms is a multi-valued and discontinuous function. Since it is difficult for a well-known multi-layer neural network to approximate such a function, a correct inverse kinematics model for the end-effectors overall position and orientation cannot be obtained by using a single neural network. In order to overcome the discontinuity of the inverse kinematics function, we propose a modular neural network system for the inverse kinematics model learning. We also propose the online learning and control method for trajectory tracking.

Inverse kinematics learning by modular architecture neural networks

Inverse kinematics computation using an artificial neural network that learns the inverse kinematics of a robot arm has been employed by many researchers. However, conventional learning methodologies do not pay enough attention to the discontinuity of the inverse kinematics system of typical robot arms with joint limits. The inverse kinematics system of the robot arms, including a human arm with a wrist joint, is a multivalued and discontinuous function. Since it is difficult for a well-known multilayer neural network to approximate such a function, a correct inverse kinematics model for the end-effectors overall position and orientation cannot be obtained by the conventional methods. In order to overcome the drawbacks of the inverse kinematics solver consisting of a single neural network, we propose a novel modular neural network architecture for the inverse kinematics model learning.

Development of Assistive Robots Using International Classification of Functioning, Disability, and Health: Concept, Applications, and Issues

Many assistive robots for elderly and disabled people have been developed in the past few decades. However, very few of them became commercially available. The major cause of the problem is that the cost-benefit ratio and the risk-benefit ratio of them are not good or not known. The evaluation of them should be done in the light of the impacts of assistive technologies on users’ whole life, both in short-term and long-term. In this paper, we propose a framework of evaluation and design of assistive robots using ICF (International Classification of Functioning, Disability, and Health). The goal of the framework is the realization of the life design and the improvement of the quality of life using assistive technologies. We describe the concept of utilizing ICF in the development process of assistive robots, and demonstrate its utility by using some examples of practical application such as the analysis of daily living, the design of assistive robots and the evaluation of assistive robots. We also show the issues of using ICF for further development of the framework.

Compact image stabilization system for small-sized humanoid

Wearable computing has been actively investigated as the dimensions of devices such as computers, sensors and motors are getting smaller. In this work our aim is to develop an image stabilization system for a small-sized humanoid robot as the first step for sharing of visual information between humans. A small-sized humanoid robot is used to simulate a person who captures visual information of a distant site. The developed image stabilization system can be used for an advanced wearable telepresence system. Requirements of an image stabilization system, which can be used as a wearable system, are compactness, lightness in weight, and real time process such as 30 frames/s. We have implemented our image stabilization system to satisfy these requirements. The developed system uses a 3D motion sensor for camera rotation detection, optical flow for camera translation detection and image process for motion compensation. The experimental results show that images from a small-sized humanoid robot can be stabilized by our developed system.

Coordinate transformation learning of hand position feedback controller by using change of hand position error norm

We need to solve the inverse kinematics problem, i.e., the coordinate transformation from the visual coordinates to the joint angle vector coordinates. In human motion control, the learning function of the hand position error feedback controller in human inverse kinematics solver is important. This paper proposes a novel model of the coordinate transformation learning of the human visual feedback controller, which uses the change of the joint angle vector and the corresponding change of the square of the hand position error norm.