COMPUTER-AIDED MECHANICAL ASSEMBLY PLANNING

Transcription

1 COMPUTER-AIDED MECHANICAL ASSEMBLY PLANNING

2 THE KLUWER INTERNATIONAL SERIES IN ENGINEERING AND COMPUTER SCIENCE ROBOTICS: VISION, MANIPULATION AND SENSORS Consulting Editor: Takeo Kanade SHADOWS AND SILHOUE77'ES IN COMPUTER VISION, S. Shafer ISBN: PERCEPTUAL ORGANIZATION AND VISUAL RECOGNITION, D. Lowe ISBN: X ROBOT DYNAMICS ALGORITHMS, F. Featherstone ISBN: THREE- DIMENSIONAL MACHINE VISION, T. Kanade (editor) ISBN: KINEMA TIC MODELING, IDENTIFICATION AND CONTROL OF ROBOT MA NIPULATORS, H.W. Stone ISBN: OBJECT RECOGNITION USING VISION AND TOUCH, P. Allen ISBN: INTEGRATION, COORDINATION AND CONTROL OF MULTI-SENSOR ROBOT SYSTEMS, H.F. Durrant-Whyte ISBN: MOTION UNDERSTANDING: Robot and Human Vision, W.N. Martin and J. K. Aggrawal (editors) ISBN: BAYESIAN MODELING OF UNCERTAINTY IN LOW-LEVEL VISION, R. Szeliski ISBN: VISION AND NA VIGATION: THE CMU NA VLAB, C. Thorpe (editor) ISBN: TASK-DIRECTED SENSOR FUSION AND PLANNING: A Computational Approach, G. D. Hager ISBN: X COMPUTER ANALYSIS OF VISUAL TEXTURES, F. Tomita and S. Tsuji ISBN: DATA FUSION FOR SENSORY INFORMATION PROCESSING SYSTEMS, J. Clark and A. Yuille ISBN: PARALLEL ARCHITECTURES AND PARALLEL ALGORITHMS FOR INTEGRATED VISION SYSTEMS, A.N. Choudhary, J. H. Patel ISBN: ROBOT MOTION PLANNING, J. Latombe ISBN: DYNAMIC ANALYSIS OF ROBOT MANIPULATORS: A Carlesian Tensor Approach, C.A Balafoutis, R.V. Patel ISBN: PERTURBATION TECHNIQUES FOR FLEXIBLE MANIPULATORS: A. Fraser and R. W. Daniel ISBN:

3 COMPUTER-AIDED MECHANICAL ASSEMBLY PIANNING edited by Luiz S. Homem de Mello Jet Pro pulsion Laboratory California Institute of Technology SukhanLee Department of Electrical Engineering Systems University of Southern Calijornia and Jet Pro pulsion Laboratory California Institute oftechnology ~. " Springer Science+Business Media, LLC

4 Library of Congress Cataloging-in-Publication Data Computer-aided mechanical assemb1y p1anning I edited by Luiz S. Homem de Mello, Sukhan Lee. p. cm. -- (The Kluwer international series in engineering and computer science: SECS 148) Includes bibliographica1 references and index. ISBN ISBN (ebook) DOI / Computer-aided engineering. 1. Assembling machines. I. Homem de Mello, Luiz S. II. Lee, Sukhan. III. Series. TA345.C dc CIP Copyright 1991 Springer Science+Business Media New York Originally published by Kluwer Academic Publishers,New York in 1991 Softcover reprint ofthe hardcover 1st edition 1991 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, mechanical, photo-copying, recording, or otherwise, without the prior written permission of the publisher, Springer Science+Business Media, LLC. Printed on acid-free paper.

5 To my parents, Ligia and Fabio Romem de Mello. To my parents, Anna and Ryung-Ki Lee.

6 Contents Foreword George A Bekey 1. Introduction 1 Luiz S. Homem de Mello and Sukhan Lee Pari I -ASSEMBLY MODELING 2. Representations for assemblies 15 Aristides A G. Requicha and Timothy W. Whalen 3. Representation of solid objects by a modular boundary model 41 Leila De Floriani, Amitava Maulik and George Nagy 4. Graphs of kinematic constraints 81 Federico Thomas 5. Relative positioning of parts in assemblies using mathematical programming 111 Joshua U. Turner Part II - ASSEMBLY PLANNING 6. Representations for assembly sequences 129 Luiz S. Homem de Mello and Arthur C. Sanderson 7. A basic algorithm for the generation of mechanical assembly sequences 163 Luiz S. Homem de Mello and Arthur C. Sanderson

7 viii 8. LEGA: a computer-aided generator of assembly plans 191 Jean-Michel Henrioud and Alain Bourjault 9. Maintaining geometric dependencies in assembly planning 217 Randall H. Wilson and Jean-FraTtfOis Rit 10. Efficiently partitioning an assembly 243 Randall H. Wilson 11. On the automatic generation of assembly plans 263 Jan D. Wolter 12. A common sense approach to assembly sequence planning 289 Richard Hoffman 13. Assembly coplanner: cooperative assembly planner based on subassembly extraction 315 Sukhan Lee and Yeong Gil Shin 14. Backward assembly planning with DFA analysis 341 SukhanLee 15. Computer aids for finding, representing, choosing amongst, and evaluating the assembly sequences of mechanical products 383 Thomas E. Abell, Guillaume P. Amblard, Daniel F. Baldwin, Thomas L. De Fazio, Man-Cheung Max Lui and Daniel E. Whitney Contributors 437 Index 439

8 Foreword Some twenty years have elapsed since the first attempts at planning were made by researchers in artificial intelligence. These early programs concentrated on the development of plans for the solution of puzzles or toy problems, like the rearrangement of stacks of blocks. These early programs provided the foundation for the work described in this book, the automatic generation of plans for industrial assembly. As one reads about the complex and sophisticated planners in the current generation, it is important to keep in mind that they are addressing real-world problems. Although these systems may become the "toy" systems of tomorrow, they are providing a solid foundation for future, more general and more advanced planning tools. As demonstrated by the papers in this book, the field of computer-aided mechanical assembly planning is maturing. It now may include: geometric descriptions of parts extracted from or compatible with CAD programs; constraints related to part interference and the use of tools; fixtures and jigs required for the assembly; the nature of connectors, matings and other relations between parts; number of turnovers required during the assembly; handling and gripping requirements for various parts; automatic identification of subassemblies. This is not an exhaustive list, but it serves to illustrate the complexity of some of the issues which are discussed in this book. Such issues must be considered in the design of the modern planners, as they produce desirable assembly sequences and precedence relations for assembly. As with other AI-based planning programs, the fundamental issues include knowledge representation and acquisition, search algorithms and inference techniques. Hence, several of the chapters of the book include discussions of model-

9 x ing and representation of parts, liaisons between them and the processes which produce them. The basic approach to problem solving is systematic search in the problem space, which immediately raises the possibility of an explosion in the number of possible solutions. Hence, many current programs include editing provisions to eliminate undesirable branches of the search tree, or prune the tree in the process of developing the plans. Yet, there are situations in which the enumeration and examination of all feasible assembly orders has advantages, as demonstrated in some of the approaches discussed here. The amount of knowledge required for the intelligent determination of feasible assembly sequences and the subsequent editing and sorting is clearly enormous. Hence, some of the methods discussed in this book include interactive features, which enable users to assist in the process by means of question answering and other computer aids. While it is evident that the automation of assembly planning is a maturing area of research, it is also evident that it is still in the research phase. Most of the authors of the papers in this book are associated with academia or with research centers. Computer-aided mechanical assembly planning is still largely an academic discipline. Its application in industry is still in the early stages. Among the reasons why its use in industry is not yet extensive are the following: 1. The computational efficiency of the today's planners limits them to assemblies with a relatively small number of parts. To handle products consisting of 60 or 80 parts, which occur in industry, current systems still need the help of a human operator. 2. The measures for the selection of desirable assembly sequences in existing planners are not yet wide enough. Cost, ease of assembly and robot gripping requirements are among the commonly used evaluation criteria. Fixturing and tooling requirements and the related issue of partial assembly turnovers are seldom discussed. Even less frequent are such practical production criteria as assembly line layout. Furthermore, many assembly processes incorporate various testing steps, which should be considered in the evaluation. Evaluation and rejection of candidate assembly sequences on the basis of such multiple criteria is clearly more difficult, but it would contribute to the perception of real world relevance of the assembly planners. Since the time associated with alternative plans and the related issues of tool change and fixture adjustment are of great industrial significance, it will also be important to integrate scheduling with planning in future systems. 3. There is a natural delay in the transfer of research knowledge from academia to industry. Industry is somewhat conservative in the introduction of new technology, which must be justified on the basis of both scientific and economic criteria. The authors of the chapters in this book report on work that addresses these problems.

10 xi I indicated earlier that assembly planning is a maturing discipline. Perhaps it is more fair to say that it has reached adolescence. It has given up the toys of childhood, but it is not quite ready to assume all the responsibilities of the rough and tumble adult world of industry. This is perhaps the most exciting time to be associated with a new technology. Readers of this book will be ready for the developments to come. George A. Bekey Computer Science Department and Center for Manufacturing and Automation Research University of Southern California Los Angeles, California

11 COMPUTER-AIDED MECHANICAL ASSEMBLY PLANNING