Distinguished Dissertations Springer London Berlin Heidelberg New York Barcelona Hong Kong Milan Paris Santa Clara Singapore Tokyo Other titles published in this Series: Extensional Constructs in Intensional Type Theory Martin Hoffman Search and Planning Under Incomplete Information: A Study Using Bridge Card Play Ian Frank Theorem Proving with the Real Numbers John Harrison Games and Full Abstraction for a Functional Metalanguage with Recursive Types Guy McCusker Hardware Evolution: Automatic Design of Electronic Circuits in Reconfigurable Hardware by Artificial Evolution Adrian Thompson Models ofS haring Graphs: A Categorical Semantics ofl et and letrec Masahito Hasegawa Chris Greenhalgh Large Scale Collaborative Virtual Environments Springer Chris Greenhalgh, BA, PhD Faculty of Science, School of Computer Science and Information Technology, University Park, University of Nottingham, Nottingham, NG7 2RD, UK Series Editor Professor c.J. van Rijsbergen Department of Computing Science, University of Glasgow, G 12 8RZ, UK ISBN 978-1-4471-1222-8 Springer-Verlag London Berlin Heidelberg British Library Cataloguing in Publication Data Greenhalgh. Chris Large scale collaborative virtual environments. - (Distinguished dissertations) 1. Virtual reality 2.Computer network architectures 3.Distinguished operating systems (Computers) 4.Multimedia systems I.Title 006 ISBN-13:978-1-4471-1222-8 e-ISBN-13:978-1-4471-0867-2 DOl: 10.1007/978-1-4471-0867-2 Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress Apart from any fair dealing for the purposes of research or private study. or criticism or review. as permitted under the Copyright. Designs and Patents Act 1988. this publication may only be reproduced. stored or transmitted. in any form or by any means. with the prior permission in writing of the publishers. or in the case of repro graphic reproduction in accordance with the terms of licences issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the publishers. © Springer-Verlag London Limited 1999 Softcover reprint of hardcover lst edition 1999 The use of registered names. trademarks etc. in this publication does not imply. even in the absence of a specific statement. that such names are exempt from the relevant laws and regulations and therefore free for general use. The publisher makes no representation. express or implied. with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility or liability for any errors or omissions that may be made. Typesetting: Camera-ready by author 34/3830-543210 Printed on acid-free paper To Vanessa, Susannah and Timothy. Preface Collaborative virtual environments (CVEs) are multi-user virtual realities which seek to actively support communication and cooperation between distributed users.These 3D audio-graphical environments have the potential to: • scale to very large numbers of simultaneous participants much more effectively than other forms of (non-spatial) computer-mediated communication such as video or audio conferencing; • provide a unified context for communication and information, for example allow ing an architectural team to have a tele-conference within a visualisation of a planned building; • maintain autonomy of individuals, and an appropriate balance of power and influ ence between participants; • support "natural" uses of space in communication, such as dynamic group forma tion, gesture and gaze direction; and • allow users to easily maintain peripheral awareness of nearby activities, and to assess the states of systems and activities at a glance. When we seek to meet these objectives we are faced with many technical and conceptual challenges. For the systems whif'll do exist, we want to know how effec tive they are and, indeed, whether our objectives are actually &chievable. Origins I have been researching, implementing, using and deploying collaborative virtual environments for just over five years, first as a post-graduate student, and more recently as a lecturer, both at the University of Nottingham. This book is my doctoral thesis, which was one of two winners of the 1998 Distinguished Dissertations in Computer Science competition. This competition is run by the Conference of Profes sors and Heads of Computing (CPHC), in conjunction with the British Computer Society (BCS), in order to select for publication a few of the best British PhDIDPhil dissertations in computer science. My thesis is presented here complete and unchanged, except for reformatting and the removal of two appendices which are not relevant for publication in this form viii Large Scale Collaborative Virtual Environments (colour images and video presentation contents). I have also added an index to help with selective reading and reference use. It is a great pleasure to see this in print, and I hope that you will find it interesting and gain insight from it. Intended Audience There are four intended audiences for this book. First, (because of its origins as a PhD dissertation) it is directly relevant to other researchers and post-graduate stu dents working in the area of multi-user virtual reality. It is particularly pertinent for researchers concerned with: • architectures for distributed virtual reality systems; • network requirements of these systems; • use and user evaluation of collaborative virtual environments; and • models of interaction and communication in space-based multi-user systems. Second, it can be used by researchers in related areas, such as single-user virtual reality, computer supported cooperative work and distributed systems, to gain a detailed understanding of the issues peculiar to systems and environments of this kind. Third, it should be accessible to any advanced computer-literate reader wishing to explore the current state of the art in specifically multi-user virtual reality. Fourth, it may serve as a model for future students (one of the objectives of the Distinguished Dissertations series) since it has been judged to be among the best cur rent dissertations in the area of computer science. The competition also requires that it combine a noteworthy contribution to the subject with a high standard of exposi tion, in such a way that it is accessible to those with other specialities. Content and Structure This book is based around two prototype collaborative virtual environment sys tems, MASSIVE-l and MASSIVE-2, which I have designed and implemented. These systems realise and extend a spatial model of interaction, which provides a conceptual and computational basis for managing communication and interaction. I have sought to combine two major perspectives: • the user's perspective within the everyday social world -considering the kinds of interaction and control which are afforded to a participant in a CVE; and • the network's perspective -considering the form and requirements of the commu nication which facilitates this interaction. Section 1.4 (page 13) explains the structure of the book and gives suggestions for reading order. The overall form (from page 15) is shown below. ix Chapter 1. Introduction Chapter 2. CVEs: a review Prototype Communication Key concept Theory Social issues overview issues Part I: Chapter 3: Chapter 4: Chapter 5: Chapter 6: Direct Original MASSIVE-l Awareness Spatial Trading relationships spatial model overview relationships Part II: Chapter 7: Chapter 8: Chapter 9: Chapter 10: Indirect Third party MASSIVE-2 Regions and Awareness driven relationships objects overview awareness communication Chapter 11. Summary, conclusions and future work Part I deals with the concepts, design, user-oriented evaluation and network requirements of MASSIVE-I, which is a unicast-based CVE implementing the spa tial model of interaction of Benford and FabIen (reviewed in chapter 3). Part II does the same for MASSlVE-2, which is a multicast-based CVE imple menting the proposed "third party object" extension to the spatial model. (See section 2.2.2 for a comparison of unicast and multicast communication.) Subsequent Developments I completed this dissertation in September, 1997. Since then work has continued on MASSIVE-2 and its successor, however thi.s cannot be integrated here because the thesis is published "as is". Perhaps the single most significant event of the past year has been the use of MASSIVE-2 for staging "Out Of This World", an experiment in Inhabited Television (see section 11.3.3). Out Of This World can be viewed as a logical extension of the Now96 event (described in section 9.2.1). Eight inhabitants (members 9f the public sitting at PCs wearing microphone-headsets) formed two teams that competed in a series of co-operative games, the winners gaining the right to escape from the doomed world. The show was projected live in a nearby auditorium to the show's viewers (also members of the public). Each team was led by a captain, played by a paid performer, who was immersed in the virtual world using a headset and tracking technology. The games were overseen by a host, who appeared in the world as a video texture. Whilst each performance took place, four virtual camera operators filmed the action. An experienced television director mixed the final show using standard outside broadcast equipment. x Large Scale Collaborative Virtual Environments Out Of This World was a great technical success, and the performances (four in total) were coherent and accessible. The inhabitants were generally involved and able to interact and collaborate. We believe that the event demonstrated the feasibility and potential effectiveness of Inhabited Television. The next key areas for development are the creation of engaging content for this format, its integration with longer-term communities of participants (at all levels -authors, performers, inhabitants and view ers) and the continuing quest to practically support much larger numbers of simulta neous participants. Acknowledgements First and foremost, I would like to thank my supervisor, Steve Benford, for providing an excellent environment in which to work and a firm conceptual and ideological foundation on which to build. Our past and present collaboration has been invaluable and most enjoyable. I would also like to thank all those members of the Communication Research Group, past and present, who have shared their time, friendship and interests. Dave Snowdon made many valuable contributions, including discussion and implementa tion. Adrian Bullock was the linchpin that made trials happen and machines work. Maarten Stenius (a visiting researcher from SICS) ported WWW-3D to MASSlVE-2. Many thanks also to Karsten Bormann, Rob Clarke, Neil Cook, Mike Craven, Mike Fraser, Tony Glover, Rob Ingram, Boriana Koleva, David Lloyd, Jim Purbrick, Mina Radenkovic, Gail Reynard, Marcus Roberts, Ian Taylor, Jolanda Tromp and Ivan Vaghi. Finally, I would like to thank Jon Crowcroft, my external examiner, for support ing my submission to the Distinguished Dissertations in Computer Science competi tion, and the panel members and other reviewers for their favourable consideration of my work. Abstract This thesis is concerned with the theory, design, realisation and evaluation of large-scale collaborative virtual environments. These are 3D audio-graphical compu ter generated environments which actively support collaboration between potentially large numbers of distributed users. The approach taken in this thesis reflects both the sociology of interpersonal communication and the management of communication in distributed systems. The first part of this thesis presents and evaluates MASSIVE-I, a virtual reality tele-conferencing system which implements the spatial model of interaction of Ben ford and Fahlen. The evaluation of MASSIVE-l has two components: a user-oriented evaluation of the system's facilities and the underlying awareness model; and a network-oriented evaluation and modelling of the communication requirements of the system with varying numbers of users. This thesis proposes the "third party object" concept as an extension to the spatial model of interaction. Third party objects can be used to represent the influence of context or environment on interaction and awareness, for example, the effects of boundaries, rooms and crowds. Third party objects can also be used to introduce and manage dynamic aggregates or abstractions within the environments (for example abstract overviews of distant crowds of participants). The third party object concept is prototyped in a second system, MASSIVE-2. MASSIVE-2 is also evaluated in two stages. The first is a user-oriented reflection on the capabilities and effectiveness of the third party concept as realised in the sys tem. The second stage of the evaluation develops a predictive model of total and per-participant network bandwidth requirements for systems of this kind. This is used to analyse a number of design decisions relating to this type of system, including the use of multicasting and the form of communication management adopted.