W.L.Chapin 3D Workshop Position Paper

Manual Interaction Techniques for 3D Computer Aided Design

Submitted as a position statement for
the Workshop on the Challenges of 3D Interaction
at CHI'94 (Boston, MA, 24-25 April 1994).
(C) 1994 Leland Stanford Jr. University

Virtual Space Exploration Lab
Center for Design Research, Stanford University
560 Panama St., Stanford, CA 94305-2232, USA

PREFACE

Due to the short notice of this workshop and its submittal date, I am writing this statement of position on behalf of my colleagues, but from distinctly my own point of view. While I will be attending CHI'94, the attendence of my colleagues is uncertain at this time. Those of us attending the conference will be demonstrating our prototype design system in the Demonstrations venue on Thursday morning, April 28th. Since there will likely be two of us presenting our work at the conference, we hope that you consider allowing both myself and my colleague attend the Challenge of 3D Interaction Workshop.

Thank you, W. Chapin.

ABSTRACT

As researchers at Stanford University's Center for Design Research, our projects have been devoted to solving communication problems in the design process. As mechanical engineers, we focus on the design to manufacture physical artifacts that serve people's needs. While part of the communication problems are between designers, some of the toughest are between designer and design model. Since most design models have a 3D representation, the designer's communication with the models is often a 3D interaction problem. One approach to assisting the design process is to center all design tools on the 3D model representation. For such an approach, the challenge of 3D interaction is central.

KEYWORDS

CAD, virtual environment, dexterous manipulation, interactive simulation, presence, spatial acoustics, manual and gestural communication, teleconference, collaboration.

INTRODUCTION

Let me begin by introducing our team members.

Larry Leifer is the principle investigator and advisor for all of our work. He has been the director of the Center for Design Research since 1984 and a full professor on the Design Division faculty in Mechanical Engineering at Stanford University since 1976.

James F. Kramer, a doctoral candidate at CDR, has focused his attention on free hand computer input and output as well as hand gesture recognition methods. He developed an instrumented glove device for interpreting hand poses for both recognition and 3D manipulation. He is also actively researching 3D displays to the hand.

Larry Edwards, a doctoral candidate at CDR, has focused his attention on 3D geometry manipulation tools. With another colleague, he developed a cut plane tool for 3D interaction in continuous perspective for conceptual shaping. He is currently finishing a thesis in conformal mapping of shapes.

Timothy Lacey, a research assistant at CDR, has focused on designer-to-designer collaboration within a 3D environment. He is pursuing research in the use of audio displays to augment 3D interaction.

I, William L. Chapin, am a post-masters research associate at CDR and have focused on dexterous manipulation of 3D objects and the creation of a testbed for a 3D model centered design process.

PROJECTS IN PROCESS

The above group has five projects in process that are relevant to the challenge of 3D interaction: TalkingGlove, CutPlane, VirtualHand, VirtualGrasp, and DesignSpace. Please refer to the attached DesignSpace summary for a brief description of these projects.

OUTSIDE EXPERIENCE

As a group, we also have considerable non-academic experience with the development of commercially available 3D interaction products. Over the past four years, a number of us have made direct or indirect contributions to the a number of 3D technology startups. Virtual Technologies, which is a technology spinoff from CDR, has produced the Virtex CyberGlove instumented glove device. Fakespace Labs, founded by a Stanford Design Division graduate, has developed interactive 3D model viewing devices. We have developed orthostereo correction algorithms and tested stereo display systems for LEEP Systems in Boston. We have worked closely with Telepresence Research, founded by Scott Fisher, formerly of the NASA VIEW project, and Brenda Laurel, editor of The Art of Human-Computer Interface Design. We have consulted with and tested products for Division, Ltd, a producer of 3D simulation systems. Crystal River Engineering, leaders in realtime spatial audio displays, have worked closely with us in the creation of 3D acoustic models and spatial audio applications. We have also consulted and product-tested for Polhemus, the originators of electro-magnetic 6D tracking instruments. This long list demonstrates the breadth and depth of our experience with design for 3D interaction, both display and input.

DesignSpace

As a project, DesignSpace encompasses much of our work in 3D interaction at CDR by its design to be a testbed for many of our ideas and theories. A summary of the DesignSpace project and system will be published in the CHI'94 proceedings and is attached to this statement for reference. There are currently three major components to DesignSpace: a collaboration facility within a shared environment that includes the design model, an interaction metaphor that places the 3D model in the user's space, and dexterous manipulation of the geometric shapes.

Since the entire DesignSpace system will be in Boston for the conference, I propose to setup the dexterous manipulation component for an interactive implementation discussion at the workshop. For the user's perspective, I would be happy to show a short video clip of what we call "spatial data presence" (see below), and present the theories and observations that led to its development.

SPATIAL DATA PRESENCE

Spatial Data Presence (SDP) is a term we have evolved from GUI to describe where we feel user interaction metaphors must evolve to provide users with usable, continuous access to all computational tools and data.

The short form of the derivation goes as follows:

The implementation of this metaphor would be in a client/server form much like X. All applications, no matter their dimensionality requirement, would run with such an interaction server. Just as one-dimensional text applications work better in a GUI, 2D applications will be more efficient in a 3D environment. The world's conception of a computer terminal must be forever changed.

DEXTEROUS MANIPULATION

One of the greatest barriers to the spread of computer tools to designers of all disciplines, is the lack of full bandwidth, dexterous manipulation. Creative people work a design with their hands under constant visualization. A keyboard and mouse take a creative designer's 60 degrees of freedom in two hands and map them into a one and two dimensional input channel respectively. We feel if we can put designers' hands on computational respresentations of forms that their creativity would be less inhibited and the computational tools will be much more effective.

We are working on instruments and algorithms that facilitate dexterous manipulation with visual, aural, and tactile feedback, or any subset.

SPATIAL AUDIO

With the rise of multimedia, GUI's are finally using more sense and senses. 3D audio for information display and input is being investigated from the beginning as an integral part of SDP. Currently, DesignSpace uses spatial audio only for human to human communication within the collaboration facilities. However, we are just initiating a project to investigate other applications of spatial audio displays and inputs for augmenting the 3D interaction.

Send Comments

_________________________________________DesignSpace CDR Stanford PaloAlto ______
chapin@cdr.stanford.edu
Last update: 18 Mar 94