Journal of the Robotics Society of Japan
Online ISSN : 1884-7145
Print ISSN : 0289-1824
ISSN-L : 0289-1824
Volume 5, Issue 2
Displaying 1-10 of 10 articles from this issue
  • Characteristics Measurements of the Alloy and Development of and Active Endoscope
    Shigeo HIROSE, Koji IKUTA, Masahiro TSUKAMOTO
    1987 Volume 5 Issue 2 Pages 87-101
    Published: April 15, 1987
    Released on J-STAGE: August 25, 2010
    JOURNAL FREE ACCESS
    Today the endoscope has become an indispensable inspection tool in the field of medical diagnosis and engineering non-destructive test. However the conventional type endoscope is not flexible enough so as to be inserted with ease inside narrow and complex path. In case of the fibersigmoidoscope, one of the typical example of the endoscope, the limited flexibility causes a great amount of pain to the patient and requires special skill to the doctor in its operation inside the lower intestine. The limited flexibility is caused by the fact that the fibersigmoidoscope has manipulatable segments only on its tip end and the rest of the stem have to follow the shape of colon by deforming itself using the force reaction from the colon wall.
    The purpose of this paper is to realize what we call an“active endoscope”, the endoscope with the stem all of which can be actively manipulated, by using Shape Memory Alloy (SMA) actuators.
    The paper first clarifies the basic characteristics of the SMA from the standpoint of its application for the servo-actuator. The simultaneous measurements of four parameters, i. e. temperature, stress, strain and electric resistance of SMA are done. The results of the SMA's four parameters are expressed in a newly introduced diagram. Based on the measurements, other diagram named temperature-stress phase are derived and the effect of the phase transformation is evaluated for the SMA's actuation performance.
    The optimal operational and annealing conditions of the SMA are made clear. The specifications of the formally proposed SMA servo actuator based on its electric resistance feedback are improved and some of the fatigue problems are solved by using these considerations.
    Based on these basic experiments and considerations, a prototype model of the active endoscope with SMA are constructed. It is 13 mm in diameter, 215 mm in length and consists of 5 active segments with array of SMA coil springs as the servo actuator under micro computer control. Each segments has 60 degree of maximum bending angle. Fiber optics are mounted at the core of the endoscope. By the joystick control of the front segments and coordination control of all the segments to shift the bending signals from tip end to the rear by the computer, the SMA active endoscope showed smooth insertion motion inside model colon path. The validity to realize the SMA active endoscope is thus demonstrated.
    Download PDF (8137K)
  • Ichiro KATO, Koichi KOGANEZAWA, Atsuo TAKANISHI
    1987 Volume 5 Issue 2 Pages 102-108
    Published: April 15, 1987
    Released on J-STAGE: August 25, 2010
    JOURNAL FREE ACCESS
    The automatic palpation system for breast cancer : WAPRO-4 was constructed. The aim of this study and development is to realize the automatic palpation instead of it by hands, and to contribute to finding breast cancer in primitive stage. The WAPRO-4 system consists with three part; the measuring instrument, the automobile instrument, and the micro-computer system. The measuring instrument has four sensory rods which independently sink a mamma by their own weight : the distances of their sinking are measured by the linear differential transformers. The industrial versatile robot (Pans Robo A 6256 C ; Matsushita electric corp.) was employed as a automobile instrument which moves the measuring instrument. The micro-computer system (PC 9801 E ; NEC corp.) control the automobile instruments, collects data from the measuring instrument during automatic palpation and performs the diagnostic procedure. The software algorithm was constructed for detecting only of tumor without affection of breathing and chest wall. The clinical testings were performed attendant with the medical doctors. The tumors of the fifteen patients among sixteen were clearly detected. The tumor of only one patient could not be detected clearly because of so small in size. These results evidently showed the effectiveness of the WAPRO-4 system and the capability of the group tests of breast cancer by using the automatic palpation system, whereas also showed that the further efforts to make the system detect smaller tumor will be needed for practical use.
    Download PDF (3826K)
  • Sadao KAWAMURA, Masashi MATSUMORI, Shigeki MATSUBAYASHI, Fumio MIYAZAK ...
    1987 Volume 5 Issue 2 Pages 109-120
    Published: April 15, 1987
    Released on J-STAGE: August 25, 2010
    JOURNAL FREE ACCESS
    It has been pointed out in much literature that when a robot performs some tasks like polishing or grinding, not only the position but also the contact force of the tools connected to the robot hand must be controlled. However, it is generally difficult for the robot to accomplish the desired position and force patterns. The main reason of this difficulty is that the reaction force and the friction force between the tools and the object which crucially affect the robot motion can not be estimated exactly, so that the input torque to generate the desired patterns can not be obtained.
    To realize the desired position and force patterns without estimating the reaction and friction forces, a new control scheme based on the iteration of trials (test motions) is proposed. In this control scheme, at the first trial the robot is excited by an input torque pattern without knowledge of the robot dynamics, reaction force, and friction force. Therefore, the robot motion does not generally coincide with the desired motion patterns. Then, at the second trial the input is simply modified by the errors which are, the differences between the real force and position patterns and the desired ones. By repeating such operations under reasonable conditions, the force and position patterns approach the desired ones. In this paper, it is theoretically shown that the real force and position patterns of robots converge to the desired ones as the trial number tends to infinity. Moreover, the effectiveness of the proposed scheme is verified by two such experiments that a serial link robot manipulator with three degrees of freedom carries out two tasks (polishing and grinding) .
    Download PDF (8605K)
  • Shigeoki HIRAI, Tomomasa SATO, Toshio MATSUSHITA
    1987 Volume 5 Issue 2 Pages 121-130
    Published: April 15, 1987
    Released on J-STAGE: August 25, 2010
    JOURNAL FREE ACCESS
    It is not easy to make a program that will cope with a variety of situations fully utilizing hands and eyes. One solution to this problem is the use of a knownledge base dealing with utilization techniques for hands and eyes.
    For this reason, task models constructed from viewpoints of the coordination of hands and eyes are essential in addition to conventional models of handling or vision function.
    This paper proposes a“Hand-Eye Cooperation Task Model” and a unified knownledge base termed the “Cooperative Task Knownledge Base” based on this task model.
    The effects of the cooperative task knowledge base are as follows. (1) The automatic, unified management of a world model in accordance with hand and eye actions, (2) The automatic use of effective vision functions for handling tasks, and (3) The automatic use of hand movements that are visually effective.
    The proposed knowledge base can be applicable to general utilization, because it covers the elementary functions for hand-eye tasks. The experiment of a valve disassembly task is performed with this knowledge base.
    Download PDF (6442K)
  • Yoshifumi KITAMURA, Masahiko YACHIDA
    1987 Volume 5 Issue 2 Pages 131-138
    Published: April 15, 1987
    Released on J-STAGE: August 25, 2010
    JOURNAL FREE ACCESS
    For recognition of 3-D shape and measurement of 3-D position of objects, it is important that a vision system can measure the 3-D data of dense points in the scene. One approach is to measure the distance on the basis of triangulation principle from the disparity of two images. This binocular vision method has, however, a difficult problem that is to find correspondence of features between two images. This correspondence problem can be solved geometrically by adding one more camera, that is, by trinocular vision.
    This paper presents the principles and implementation details of this trinocular vision. Based on the proposed method, we have made several experiments. By these experiments, it is found that many correct correspondence can be established even for images of complex scene by only geometrical constraint of the trinocular vision. However, when there are dense edge points in the image, multiple candidate points are found and unique correspondence cannot be established. To solve this problem, we will discuss two approaches in this paper.
    Download PDF (4563K)
  • Munehisa TAKEDA, Takahiro MASUDA, Akemi FUTAKAWA, Sadao KAWAMURA, Fumi ...
    1987 Volume 5 Issue 2 Pages 139-149
    Published: April 15, 1987
    Released on J-STAGE: August 25, 2010
    JOURNAL FREE ACCESS
    Learning control based on the repetitive operation of robots is one of the useful methods to realize high speed and high precision control for robots. It betters the next operation of a robot by using previous operation's data. From the practical viewpoint, the trial number of learning should be small. However, most previous studies forcuseed on the proof of learning convergence, and few study considered about the convergence rate for learning control.
    In this paper, we propose a learning control scheme with high convergence rate. This proposed learning control scheme takes advantage of linear approximation of the inverse robot dynamics. In this learning algorithm, modified control inputs are generated by using position, velocity, and acceleration errors, multiplied position servo gains, velocity servo gains, and motor inertia, respectively. This method is very practical because it does not necessitate the identification of phisical parameters of a robot. Some simulation results are given in order to examine the applicability of this proposed learning scheme to various kinds of industrial robots.
    Download PDF (2957K)
  • Koichi OSUKA, Hirokazu MAYEDA
    1987 Volume 5 Issue 2 Pages 150-157
    Published: April 15, 1987
    Released on J-STAGE: August 25, 2010
    JOURNAL FREE ACCESS
    For improving the ability of manipulators or developing the CAD systems, it is necessary to develop an identification method for manipulators and an effective computational method for inverse dynamics problem.
    We have already developed a practical identification method for serial manipulators, and there is an effective algorithm for inverse dynamics problem proposed by J. Y. S. Luh. But, there are some differences between the parameters set suited for our method and the parameters set suited for Luh's algorithm.
    For that reason, we had to treat the above problems as different problems.
    In this paper, we introduce the concept of virtual parameters in order to treat both problems in a unified manner. First, we discuss the parameter form suited for each problem. Next, we define the virtual parameters using the results of the identification problem to solve the inverse dynamics problem.
    Download PDF (1913K)
  • Yoshio HAYASHI
    1987 Volume 5 Issue 2 Pages 158-162
    Published: April 15, 1987
    Released on J-STAGE: August 25, 2010
    JOURNAL FREE ACCESS
    Download PDF (1284K)
  • Fumio HARASHIMA
    1987 Volume 5 Issue 2 Pages 167
    Published: April 15, 1987
    Released on J-STAGE: August 25, 2010
    JOURNAL FREE ACCESS
    Download PDF (112K)
  • Koichi SUGIMOTO
    1987 Volume 5 Issue 2 Pages 168
    Published: April 15, 1987
    Released on J-STAGE: August 25, 2010
    JOURNAL FREE ACCESS
    Download PDF (130K)
feedback
Top