Purpose
The Virtual Environments Technical Group is concerned with human factors issues associated with human-virtual environment interaction. These issues include maximizing human performance efficiency in virtual environments, minimizing health and safety problems, and circumventing potential social issues through proactive assessment.
Joining
To join the Virtual Environments Technical Group please contact the Human Factors and Ergonomics Society, P.O. Box 1369, Santa Monica, CA 90406, U.S.A. Dues is $3 per year. You do not have to be a member of the Human Factors and Ergonomics Society to join. You can be placed on our email mailing list (at no charge) by sending your email address to: mourant@coe.neu.edu.
VETG Officers - 1996-1997
(Elections for new officers will be held at the 41st Annual Meeting.)
The present officers of the Virtual Environments Technical Group are:
Chair: Ronald R. Mourant
Department of Mechanical, Industrial and Manufacturing Engineering
Northeastern University, Boston, MA 02115
Phone: 617 373-3931 mourant@coe.neu.edu
Annual Meeting Program Chair: Kay M. Stanney
Department of Industrial Engineering and Management Systems
University of Central Florida, Orlando, FL 32816
Phone: 407 823-5582 stanney@iems.engr.ucf.edu
41st Annual Meeting
The Human Factors and Ergonomics Society's 41th Annual Meeting will be held
September 22-26, 1997 in Albuquerque, New Mexico, U.S.A.. The Virtual Environments
Technical Group is sponsoring technical paper sessions and poster presentations.
Abstracts are presented below. For further information on specific papers
please contact the authors directly. Papers will be published in the Proceedings
of the 41st Annual Meeting.
Wednesday, September 24, 8:30 A.M.
Training in Virtual Environments
ASSESSING VIRTUAL REALITY’S POTENTIAL FOR TEACHING ABSTRACT SCIENCE
Marilyn Salzman, Chris Dede, Debra Srague
George Mason University
msalzman@gmu.edu
R. Bowen Loftin
NASA & University of Houston
Understanding how to leverage the features of immersive, three-dimensional (3-D) virtual reality to meet user needs presents a challenge for human factors researchers. This proposal describes our approach to evaluating this medium's potential as a tool for teaching abstract science. It describes some of our early research outcomes and discusses an evaluation comparing a 3-D VR microworld to an alternative 2-D computer-based microworld. Both are simulations in which students learn about electrostatics. The outcomes of the comparison study suggest: 1) the immersive 3-D VR microworld facilitated conceptual and three-dimensional learning that the 2-D computer microworld did not, and 2) VR's multisensory information aided students who found the electrostatics concepts challenging. As a whole, our research suggests that VR's representational abilities have promise for teaching and for visualization and that characteristics of the learning experience such as usability, motivation, and simulator sickness are important part of assessing this medium's potential.
LANDMARK LOCATION AND DISTANCE ESTIMATION IN A LARGE SCALE VIRTUAL ENVIRONMENT
Robert C. Allen and Daniel P. McDonald
University of Central Florida, Orlando, Florida
allenr@stricom.army.mil
Michael J. Singer
U.S. Army Research Institute for the Behavioral and Social Sciences
The U. S. Army Research Institute (ARI), under the auspices of the Department
of Defense, is currently investigating the usefulness of Virtual Environment
(VE) technology as a training tool for the dismounted soldier. For example,
Singer, Allen, McDonald, and Gildea (1997) conducted an experiment that
examined spatial knowledge acquisition after exposure to a VE.
Participants were required to traverse a 400 m route, in 100 m increments,
through a large scale (2000 m x 2000 m) natural VE. Two groups that trained
in the VE (Low, Hi) were compared to a control group (Topographical Map
study). The Hi VE group traversed the terrain by walking on a treadmill;
the Low VE group used a joystick to move through the environment. Two VE
terrains were constructed. Terrain One was an abstract terrain that had
several distinctive landmarks, e.g., road cut, a tall, pointed hill, etc.
while Terrain Two was modeled from an aerial photo of a portion of Fort
Benning, GA. Terrain Two's landmarks were less distinctive, i.e., flatter,
than were the landmarks of Terrain One. The current paper reexamined the
Singer et al. (In Press) data, focusing on distance estimates made by participants
during the training phase of the study. Results indicate that people's ability
to locate landmarks within a large scale VE improves with experience. However,
experience in the environment did not seem to affect distance estimation
accuracy. It was also shown that participants were more accurate in terms
of their ability to locate proximal, as opposed to distal, landmarks. However
landmark proximity did not influence distance estimation accuracy. Contrary
to our predictions, more accurate distance estimations were made by the
Low VE group. This result is explained in terms of increased task demands.
A USABILITY ASSESSMENT OF VIRTUAL REALITY SIMULATION FOR AERIAL GUNNER TRAINING
Denise Silverman, Hughes Training, Inc.
silverdmtb@aol.com
Alan Spiker, Anacapa Sciences, Inc.
The 58th Training Support Squadron (58 TRSS) has begun use of a one-of-a-kind virtual realitv (VR) simulator for training aerial gunner/scanner (AG\S) tasks. This paper briefly describes the development considerations of the device and an initial usability assessment of the helmet mounted display (HMD) and the helmet. Eleven experienced rotary-wing instructors with varying levels of experience in the Aerial Gunner Scanner Simulator (AGSS) were surveyed. Their ratings and comments regarding 17 helmet components are summarized. The results suggested the need for improved CRTs and faster update rates for the head trackers. The results also indicated that helmet comfort factors and fitting procedures (e.g., alignment) were problematic. Several recommendations are made based on these central findings to AGSS-users in particular and VR-users in general.
STATIC ANTHROPOMETRIC VALIDATION OF "DEPTH v.4.1"
Kristie Nemeth, Miami University, Oxford, Ohio
nemeth@iglou.com
The current project examines the validity of the human model in the Design and Evaluation of Personnel, Training and Human Factors (DEPTH) application. A total of 28 anthropometric measurements taken from 17 human volunteers were compared to the measurements taken from the corresponding human model in DEPTH v4.1 which is based on the JACK" model. Although a few dimensions were accurately represented, many had large discrepancies. Some of the measurement deviations could be explained by different measurement procedures, but this cannot account for all of the error. Given the large discrepancy found in the hand and forearm sections of the body, the current version of DEPTH v4.1 would not allow a designer to accurately simulate reach and grasping tasks. Future research should continue to consider additional anthropometric dimensions which are necessary to realistically simulate a human figure. In addition to static body measurements, it is necessary to examine the human body in motion. To accurately simulate a human performing a task, information about the size, shape and movement of the model are necessary.
Thursday, September 25, 3:30 P.M.
Empirical Studies in Virtual and Augmented Reality
EMPIRICAL MODELING OF PERCEIVED PRESENCE IN VIRTUAL ENVIRONMENTS USING SEQUENTIAL EXPERIMENTATION TECHNIQUES
Michael Snow
Wright Laboratory
snowmp@wl.wpafb.af.mil
Robert Williges
VPI & SU, Blacksburg, Virginia
williges@vt.edu
Investigation of perceived presence in virtual environments has so far
been
limited. This is due in part to the lack of quantitative, ratio-scale measures
of this phenomenon. Another limitation is that research in this area has
consisted largely of single studies examining, at most, a few independent
variables at once. A series of studies was conducted to test a ratio-scale
measure of presence and to examine the effects of a large number of VR system
parameters on this measure. This paper reports the results of the effort
to
examine simultaneously the main effects, second-order effects, and two-way
interaction effects of twelve variables on perception of presence in a virtual
environment. This was done using sequential experimentation techniques.
The
consequent effort to empirically model these effects using polynomial regression
is also described. The procedures and limitations of sequential
experimentation, its application to examination of perceived presence, and
caveats associated with data bridging and empirical modeling of the results
are
discussed. The final empirical model that was arrived at is presented and
its
implications for the use of virtual environments and virtual environment
technology are also discussed. It is concluded that sequential experimentation
is a useful tool for examining perceived presence, but that the subjective
nature of this phenomenon and large apparent individual differences can
make
data bridging problematic.
SCIENTIFIC VISUALIZATION PERFORMANCE IN VIRTUAL ENVIRONMENTS
Edward P. McCormick
U.S. Air Force, Edwards AFB
MCCORMIC@BFTF.EDW
Christopher D. Wickens, Michelle Yeh, and Rachel Banks, University of
Illinois at Urbana-Champaign
cwickens@psych.uiuc.edu
Three frames of reference (full exocentric, full egocentric and tethered) were contrasted in three experiments to measure performance of four different scientific visualization subtasks: search, travel, local judgment support and global judgment support. Participants were instructed to locate and follow a designated path through simple virtual environments and to answer simple questions about that environment. Each subject completed 5 trials in every frame of reference condition (exocentric, egocentric and tethered) with the subject's first trial serving as a practice trial. The results revealed that frames of reference that utilize egocentric or tethered viewpoints support better travel performance, especially when nearing the target. However the exocentric frame of reference supported better Performance in the search subtasks and possibly the local judgment, and global judgment subtasks.
INFLUENCE OF HEAD MOTION ON THE JUDGED DISTANCE OF VIRTUAL OBJECTS
Stephen R. Ellis, Brian M. Menges, Richard Jacoby, Bernard D. Adelstein
and Jeffrey W. McCandless
NASA Ames Research Center, Moffett Field, CA
silly@eos.arc.nasa.gov
Head motion is shown to improve the judged distance to monocularly viewed, space stabilized virtual objects presented via head-mounted see through displays. In contrast to similar judgments made with stereoscopic displays, the resulting judgments are not veridical, have considerable variability, and some of their frequency distributions markedly deviate from normality. Comparison of distance judgment to monocularly viewed real objects suggests that the observer's judgment difficulties arise not from the task itself but imperfections in the displayed imagery. Future work will examine full system display latency as a cause of the imperfect depth rendering observed in our current displays.
PERCEPTUAL EFFECTS IN ALIGNING VIRTUAL AND REAL OBJECTS IN AUGMENTED REALITY DISPLAYS
Paul Milgram and David Drascic
University of Toronto, Toronto, CA
milgram@ie.utoronto.ca
The concept of Augmented Reality (AR) displays is defined, in relation to the amount of real (unmodelled) and virtual (modeled) data in an image. AR encompasses those displays in which real images, such as video, are enhanced with computer generated graphics. AR is seeing increased use, inter alia, in telerobotics, medicine and design. For the important class of stereoscopic AR displays, several factors may cause potential perceptual ambiguities, however, which manifest themselves in terms of decreased accuracy and precision whenever virtual objects must be aligned with real ones. Using the ARGOS Augmented Reality through Graphic Overlays on Stereo-video) system, a Virtual Tape Measure (VTM) has been developed, for performing quantitative 3D measurements on real-world stereo images. An experiment was conducted to assess both the magnitude of the above mentioned perceptual effects and the effectiveness of a computer assisted VTM. Results showed comparable accuracy for the two, but better precision for the assisted VTM.
Friday, September 26, 8:30 A.M.
Virtual Environment Performance: The Implications of Fitts’ Law
FIELD OF VIEW EFFECTS ON A DIRECT MANIPULATION TASK IN A VIRTUAL ENVIRONMENT
William P. Janson and Kenneth A. Aldrich
Logican Technical Services, Inc.
bjanson@al.wpafb.af.mil
Robert G. Eggleston, Human Engineering
Division, Armstrong Laboratory
It is important for designers of virtual reality (VR) systems to understand how properties of the system interact with human processes to impact overall system performance. The following experiments investigated the field-of-view (FOV) characteristics of a VR system and how they affected aimed movement behavior across a natural and virtual environment. Utilizing three levels of FOV, and two levels of task difficulty, subjects performed a reciprocal tapping task. FOV was found to differentially affect movement time performance between the two environments for the moderate level of task difficulty, but not for the more highly difficult level. This significant effect for longer movement times in the virtual environment was traced to the smallest FOV, involving those targets appearing outside the instantaneous FOV. Movement time differences between environments are also discussed in terms of stereopsis and helmet weight considerations.
INVESTIGATING COORDINATION IN MULTIDEGREE OF FREEDOM CONTROL
Shumin Zhai
IBM Almaden Research Center, San Jose, CA
zhai@almaden.ibm.com
In these two companion papers, methods developed in a series of studies in the 1940's and 1950's are applied to the analysis of 6 DOF control devices used in modem human machine systems such as virtual environments. Contrary to the early studies, the current work showed that the simultaneous time on target in multidegree of freedom tracking was higher than the product of component time on target scores. The distribution of linear correlation coefficients between the tracking errors of different degrees of freedom tended to be skewed towards the positive values. These results suggested that subjects discoordination in early multidegree of freedom tracking studies was likely due the limitation of human machine interfaces at that time. With well designed interfaces, subjects exhibited more coordinated trials than discoordinated trials in multidegree of freedom tracking.
FITTS AND VR: EVALUATING DISPLAY AND INPUT DEVICES
Mike McGee, Brian Amento, Patrick Brooks, Hope G. Harley, VPI & SU,
Blacksburg, Virginia
mmcgee@vt.edu
This paper describes an experiment using Fitts' Law to evaluate performance in target acquisition tasks using typical virtual reality (VR) and computer workstation displays and input devices. A VR4 helmet-mounted display (HMD), a 17 inch CRT monitor, a space mouse, and a regular computer mouse were tested with various parameters of classic Fitts' Law experiments: width of target, distance of target from starting point, angle of target from starting point(theta). Movement times were collected in a standard 2D target acquisition task. Input device, distance from target, width of target, and theta were significant. Display type did not result in significantly different movement times. The implications of using Fitts' Law as an evaluation tool for VR display and input devices are discussed, including extensions of the task into three dimensions.
VIRTUAL VISUAL AND AUDITORY DISPLAY AIDS FOR A PEG INSERTION TASK
Kenneth Nemire
Interface Technologies, Capitola, California
knemire@netcom.com
Inadequate depth and distance cues in many commercial head-mounted displays make it difficult for users to know when they have touched a virtual object with their virtual cursor. Studies of teleoperation tasks show that using stereoscopic displays improves performance, especially in situations with increased scene complexity and decreased object visibility. However, many aspects of these studies prevent generalization of the results to virtual environment systems. In this paper, the effects of various visual and auditory display enhancements were evaluated to determine whether they may replace depth information provided by stereoscopic visual displays. A peg insertion task provided a common test situation in which to compare various enhancements. Participants wore a commercial head-mounted display and spatial trackers on the head and hand. Results indicated conditions under which visual and auditory enhancements to monocular displays resulted in performance that was not different from using stereoscopic displays. Implications of the results are discussed.
Friday, September 26, 10:30 A.M.
Spatial Awareness in Virtual Environments
MEASURING ATTENTIONAL DEMAND IN A VIRTUAL ENVIRONMENTS DRIVING SIMULATOR
Ronald R. Mourant and Zhuoyun Ge
Northeastern University, Boston, Massachusetts
mourant@coe.neu.edu
The measurement of drivers’ attentional demand is of importance to highway designers as well as to human factors engineers concerned with in-vehicle navigational displays and their user interface. When attentional demand is great, older drivers in particular are subjected to information overload which may adversely affect their driving performance. In 1967, Senders, et. al. measured drivers’ attentional demand using a visual occlusion device. They found that attentional demand increased with vehicle velocity, but their study was limited to off-road driving due to safety considerations. The present study focused on the establishment of baseline quantitative measures of attentional demand as a function of vehicle velocity, road curvature, and amount of on-coming traffic. Eight experienced drivers drove a virtual environments driving simulator while controlling the visibility of the forward scene via a foot switch. They were instructed to stay in their lane and obey all traffic regulations, while trying to keep the forward scene visible as little as possible. Attentional demand was calculated as the percent of time drivers viewed the forward scene during a run. While driving at 100 kph the average attentional demand (86%) was significantly higher (p < .01) then when driving at 33 kph (77%). The finding that an increase in vehicle velocity of 67 kph resulted in only a 9% increase in attentional demand was somewhat surprising. As expected, the attentional demand of drivers on curves (85%) was significantly higher (p < .05) then when driving on a straight road (81%). The effect of traffic versus no traffic on curves was also statistically significant. The attentional demand on curves with moderate on-coming traffic was 88% versus 80% when there was no traffic. This study demonstrated the feasibility of using a virtual environments simulator to measure drivers’ attentional demand. Future research will study attentional demand using novice and older drivers.
AN EXPERIMENTAL COMPARISON OF THREE METHODS FOR COLLISION HANDLING IN VIRTUAL ENVIRONMENTS
Jeffrey Jacobson and Michael Lewis
University of Pittsburgh, Pittsburgh, PA
goshen@cmu.edu
This study compares three common strategies for handling collisions between the user's virtual body and other objects in a cluttered virtual environment. Test subjects seek "treasures" in a maze of narrow corridors which are embedded in a jumble of irrelevant shapes. The application ran on a PC, with the mouse and screen as the interface. The user can either pass through an object, is forced to stop completely, or is deflected around it. Data show that the third strategy best facilitates goal-seeking behavior with this interface and for this type of problem. This result is significant, because collision handling is critically important to the usefulness of Virtual Reality applications. Furthermore, the screen and-mouse is both the most common and least studied interface for virtual environments.
FACTORS INFLUENCING SPATIAL AWARENESS AND ORIENTATION IN DESKTOP VIRTUAL ENVIRONMENTS
Dennis Neale
VPI & SU, Blacksburg, Virginia
dneale@vt.edu
This study investigated influences on human spatial orientation and representation in a perspective display as the result of manipulating the geometric field of view (GFOV) and incorporating symbolic virtual environment enhancements in the interface based on visual momentum (VM) techniques. Sixty participants, pretested for spatial ability, were required to navigate through a virtual office building while performing a variety spatial orientation tasks. A 3 x 2 x 2 mixed-subjects design compared three levels of GFOV, two levels of VM, and two levels of Difficulty. The findings indicated that decreases in GFOV produced spatial orientation and representation errors, and as task difficult increased, errors were more pronounced. Furthermore, VM was shown to alleviate errors in spatial orientation and representation, especially when task difficulty was increased. Design recommendations are discussed based on the implications of the results.
USING A COMPUTER SIMULATED WORLD TO STUDY BEHAVIORAL COMPLIANCE WITH WARNINGS: EFFECT OF SALIENCE
Barbara Glover and Michael S.Wogalter
North Carolina State University, Raleigh, NC
Barbara@poe.coe.ncsu.edu
Warning signs and consumer product labels are intended to alert persons to potential dangers. Despite its importance, empirical research measuring behavioral effects of warnings is limited due to the difficulties of conducting behavioral compliance studies. One cannot ethically expose a person to truly hazardous situations. This study used a computer simulated world as a new method for studying behavioral compliance. Such simulations can be constructed to appear realistic, thus maintaining ecological validity, while allowing the experimenter control over experimental conditions. Three warning factors, time stress, salience, and type of sign, were manipulated to determine their effects on exiting from an underground mine. Results indicated high salience increased compliance for all participants, while the time stress and type of sign manipulations failed to produce any significant effects. In addition, more women than men complied with warnings. It appears that virtual simulated worlds may be a promising new methodology for studying behavioral compliance with warnings.
Progress is being made on the necessary hardware and software needed for virtual environments research. The items below should help subjects realize a better sense of immersion when using a virtual environment.
Virtual Research Systems, Inc. continues to make high quality ergonomically
designed HMDs. Their V6 model has VGA resolution (640 x 480) and weights
only 29 ounces. Their latest model, the V8, has triple the horizontal resolution
of the V6. Phone: 408-748-8712
http://www.virtualresearch.com
InterSense, Inc. has precision motion trackers for use with HMDs and
gloves. These trackers use inertial sensing technology and eliminate jitter
common to other systems. Phone: 617 499-0020
http://www.isense.com
SensAble Technologies makes haptic interface devices. Their PHaNToM systems
work by having the user insert his or her finger into a thimble. The device
interprets the finger’s position in 3D and applies an appropriate and
variable resisting force. Smooth or textured surfaces, complex curves and
sharp corners, solid or complaint objects and friction can all be accurately
represented to the user via the PHaNToM. Phone:617 621-0150
http://www.sensable.com
CALENDAR
41st Annual Meeting of the Human Factors and Ergonomics Society, Albuquerque,
New Mexico, September 22-26, 1997. Phone: 310 394-1811
http://www.hfes.vt.edu/HFES