ASSESSING SIMULATOR SICKNESS IN A SEE-THROUGH HMD: EFFECTS OF TIME DELAY, TIME ON TASK, AND TASK COMPLEXITY W. Todd Nelson Robert S. Bolia Research Psychologist Computer Scientist Air Force Research Laboratory Air Force Research Laboratory Wright-Patterson AFB, OH Wright-Patterson AFB, OH Merry M. Roe Rebecca M. Morley Human Factors Psychologist Department of Psychology Sytronics Inc. Clemson University Dayton, OH Clemson, SC Abstract is intuitively appealing, and many potential applica- tions have been suggested, including: (1) design, manu- Advances in helmet-mounted displays (HMDs) facturing, and marketing; (2) medicine and health care; have permitted the design of “see-through” displays in (3) teleoperation for hazardous operations; (4) training; which virtual imagery may be superimposed upon real (5) education; (6) information visualization; (7) tele- visual environments. Such displays have numerous communication and teletravel; (8) entertainment and potential applications; however, their promise to im- art, and (9) national defense (Durlach & Mavor, 1995). prove human perception and performance in complex In the case of medical applications, the treatment task environments is threatened by numerous techno- of tumors by radiation serves as a striking example of logical challenges. Moreover, users of HMDs may be how see-through displays may significantly enhance the vulnerable to symptoms associated with simulator sick- quality and safety of medical procedures. In short, the ness. The primary objective of this investigation was to goal of radiation treatment is to deliver a high dosage assess subjective ratings of simulator sickness as a of concentrated radiation to the tumor, while at the function of time delay, time on task, and task complex- same time minimizing the radiation exposure and dam- ity. Participants attempted to center a reticle over a age to healthy tissue. However, the patient- and tumor- moving circular target using a see-through HMD while specific nature of radiation treatment necessitates a concurrently performing a visual monitoring task dis- methodology that provides exceptional precision. One played on a computer monitor. Results indicated that possibility would be to combine medical imaging tech- simulator sickness ratings varied directly with time on nology with see-though visual displays, thus enabling task, while performance efficiency and ratings of per- the surgeon to choose the optimal path for the radiation ceived mental workload were not mediated by this fac- beams and thereby maximizing the dosage of radiation tor. Furthermore, the time delay manipulation that af- to the tumor while minimizing radiation damage to fected performance efficiency and operator workload healthy tissue and organs (see Rheingold, 1991). did not generally influence SSQ ratings. These find- A further advantage of augmented display technol- ings are discussed in terms of their implications for ogy is that it may enable surgeons to practice and re- practical implementation of see-through HMDs in hearse surgical procedures, including alternative plans multi-task environments. of action in the event of unexpected complications. Again, the idea is to combine see-through HMD dis- Introduction plays with information provided by x-rays, magnetic resonance imaging (MRI) scans, and computer tomog- As described by Durlach and Mavor (1995), ad- raphy (CT) scans so as to enhance the capability of vances in helmet-mounted display (HMD) technology surgeons in the planning and completion of surgical have permitted the design of display systems which procedures. combine virtual and real environments. Such sytems With regard to the application of see-through make use of “see-through” HMDs, which allow virtual HMDs for tactical aviation, numerous researchers imagery to be superimposed upon real visual envi- (Adam, 1994; Beal & Sweetman, 1994; Furness, 1986; ronments. Indeed, the notion of using see-through Wells & Griffin, 1987) have noted the potential tactical HMDs to augment visually-complex real environments advantages. For example, see-through HMDs are ca- _____________________________ pable of displaying flight-critical information irrespec- Presented at the IMAGE 2000 Conference tive of the pilot’s line of gaze, which is consistent with Scottsdale, Arizona 10-14 July 2000. the suggestion of Stinnett (1989) that the ability to “look-around” is advantageous, if not crucial, when Report Documentation Page Form Approved OMB No. 0704-0188 Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. 1. REPORT DATE 2. REPORT TYPE 3. DATES COVERED 2000 N/A - 4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER Assessing Simulator Sickness in a Self-Through HMD: Effects of Time 5b. GRANT NUMBER Delay, Time on Task, and Task Complexity 5c. PROGRAM ELEMENT NUMBER 6. AUTHOR(S) 5d. PROJECT NUMBER 5e. TASK NUMBER 5f. WORK UNIT NUMBER 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION Air Force Research Laboratory Wright-Patterson AFB, OH 45433 REPORT NUMBER 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S) 11. SPONSOR/MONITOR’S REPORT NUMBER(S) 12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release, distribution unlimited 13. SUPPLEMENTARY NOTES The original document contains color images. 14. ABSTRACT 15. SUBJECT TERMS 16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF ABSTRACT OF PAGES RESPONSIBLE PERSON a. REPORT b. ABSTRACT c. THIS PAGE UU 7 unclassified unclassified unclassified Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18 performing low altitude, terrain-avoidance maneuvers. eyestrain, headache, difficulty focusing and blurred In addition, see-though display technology may afford vision ... Aftereffects associated with simulator sick- ness include postural instability, dizziness, and all-weather, 24-hour flight operations. Finally, when flashbacks. Flashbacks, which include illusory sen- see-though HMDs are used in combination with tar- sations of climbing and turning, sensations of nega- geting-displays, pilots gain the capability to track and tive g, and perceived inversions of the visual field, designate targets, as well as aim and guide weapons, by are particularly problematic because of their sudden unexpected onset and risk to safety. (p. 62) line of gaze. The advantages associated with see-though HMDs It is important to point out that while the symptomotol- may also extend to the design of effective human- ogy of motion sickness and simulator sickness overlap, machine interfaces for uninhabited aerial vehicles the pathognomonic signs of the former (i.e., vomiting (UAVs). As pointed out in a recent report on UAVs by and retching) are infrequent in the latter (Kennedy, the United States Air Force Scientific Advisory Board, Lane, Lilienthal, Berbaum, & Hettinger, 1992). Yet, as “the human’s flexibility and capability for inductive noted by Kennedy and his colleagues, the potential for reasoning are desirable attributes that justify the reten- negative aftereffects is one of the most serious prob- tion of a significant supervisory and intervention capa- lems associated with simulator sickness. Aftereffects bility during UAV operations in the foreseeable future” associated with flight simulation have included disrup- (Worch, 1996, p. 7-2). The latter part of this recom- tions of postural control (Kennedy, Fowlkes, & Lilien- mendation implies that there may be situations in which thal, 1993; Kennedy & Stanney, 1996), the illusions of UAV operators are required to assume manual control flying and rotating, and the perceived inversion of the of certain UAV functions, such as automatic targeting visual field (Kennedy et al., 1992). functions and flying/landing the UAV. In these cases, Besides the untoward effects associated with augmented displays may permit UAV ground station simulator sickness, Kennedy et al. (1992) have sug- operators to perform the additional manual control gested that its occurrence may drastically inhibit train- tasks while concurrently performing their normal ing effectiveness. For example, in order to reduce ground station monitoring tasks. Moreover, similar to symptomatology, operators wearing HMDs may adopt the medical applications, see-through displays may pro- behavioral strategies that are inappropriate for the task vide UAV ground station operators an effective means at hand, such as closing their eyes, restricting head for reviewing and rehearsing mission scenarios while movement, or looking away from vection-inducing vis- the UAV fleet is en route to various tactical destina- ual displays (Kennedy et al., 1992). Moreover, be- tions. haviors acquired to reduce symptomatology may also Despite these advantages, see-through displays are jeopardize the positive transfer of skills to other VEs or confronted by many technological challenges, including real environments. Paradoxically, extensive training or misalignment of virtual imagery with real world objects mission rehearsal with HMDs in which simulator sick- (Azuma & Bishop, 1994), optical distortion and glare, ness is prevalent may actually impair one’s ability to and problems generic to most HMDs (e.g., helmet fit perform these tasks in the real world. and discomfort, field of view limitations, suboptimal Hettinger and Riccio (1992) have noted that mani- resolution, and issues involving time delay). Further- festations of simulator sickness often occur in the pres- more, while these technical limitations have been ence of excessive time delays, which may cause virtual shown to adversely affect performance efficiency, con- images to appear to float or swim-around in the HMD, trol strategies, and operator workload, it is also likely an effect that has been described as subjectively dis- that they occasion the onset of simulation sickness turbing and nauseogenic (Azuma & Bishop, 1994; (Kennedy, Lanham, Drexler, Massey, & Lilienthal, Rheingold, 1991; So & Griffin, 1991). While anecdotal 1995). Consistent with this view, Durlach and Mavor evidence seems to support the notion that visual time (1995) have noted that HMDs increase the likelihood delays in HMDs play an important role in the occur- that motion sickness will be a significant problem in rence of simulator sickness, research in this area has synthetic environments, potentially compromising op- been sparse. In a recent study addressing the effects of erator safety and acceptance, and thus mission effec- time delay on simulator sickness in a non-see-through tiveness. HMD, Draper and his colleagues (in review) concluded that reports of simulator sickness did not vary as a Simulator Sickness function of time delay. These results are consistent with those of Nelson (1996), who found that overall As defined by Kennedy, Allgood, and Lilienthal levels of simulator sickness remained unchanged as (1989), simulator sickness refers to time delay was increased in a tracking task performed by operators wearing an HMD. Collectively, these motion sickness-like symptoms that occur in aircrew studies indicate that time delay may be a necessary but during and following training. Symptoms include general discomfort, stomach awareness, nausea, dis- not sufficient condition for the development of simula- orientation and fatigue. There is also a prominent tor sickness. component of visually related disturbances such as The purpose of the present experiment was to as- the visual monitoring task. Trials 11-20 involved both sess the effects of time delay, time on task, and task the tracking and monitoring tasks. Prior to the initia- complexity on the incidence of simulator sickness in a tion of the main experimental sessions, all participants task performed using a see-through HMD. This extends completed five 5-min practice trials of the secondary to a more operationally relevant environment the stud- visual monitoring task. The purpose of the practice ies by Draper et al. (in review) and Nelson (1996), trials was to acquaint participants with the response which were limited by their use of single task environ- procedures for the task and to ensure that they were ments and non-see-through HMDs. The decision to able to perform the task at ceiling level. employ a see-through HMD was motivated by the fact Participants used a Kaiser Electronics SimEye that one of their principal advantages is the ability to 2500 HMD to track a moving visual target. The support operators in multi-task environments, such as SimEye 2500 HMD employs an optical relay system to the tracking of virtual imagery while concurrently transfer video images from a pair of green phosphor monitoring events in the real world. Furthermore, monochrome cathode ray tubes (CRTs) to the partici- Azuma and Bishop (1994) have noted that when see- pant's eyes. It features a high resolution (1280 x 1024 through displays are used for this purpose in time- pixels) binocular display and was configured to provide delayed augmented realities, virtual objects appear to subjects with a 60° (horizontal) × 40° (vertical) field of “swim around” real objects, a condition which may view (FOV). The optical focus range of the SimEye provoke a symptomatology consistent with simulator 2500 extends from 3.5 feet to infinity, and was set to sickness. These effects may be exacerbated by in- infinity in the present experiment. The SimEye 2500 creased levels of time delay in conjunction with a visu- weighs approximately four pounds and was configured ally-complex real environment. as a see-through display, thereby allowing participants to view the visual display on which the monitoring task was presented. Method The head-slaved tracking task employed target motion patterns (see Fig. 1), or forcing functions, that Participants consisted of the sum of three sine waves with funda- mental frequencies of 0.067, 0.117, and 0.233 Hz in Seven naïve participants, 3 male and 4 female, azimuth, and 0.083, 0.167, and 0.217 Hz in elevation. served in the experiment. Their ages ranged from 20 to Target motions were restricted to ±30° in azimuth, and 32 years with a mean of 24.35 years. Participants re- ±20° in elevation. Different target motions were gener- ported normal or corrected-to-normal vision, and indi- ated for each trial by randomly assigning phase values cated that they were not highly susceptible to motion at each of the three fundamental frequencies in azimuth sickness. In addition, all participants reported no prior and elevation. experience with head-slaved tracking tasks using a see- through HMD. Individuals were paid for their partici- 40 pation. 30 Experimental Design 20 tawndo nAtiom mwei indthaeilln a+-ys u1cb0oj0ne dcmittssi o)d nwesse i(rgneno cmwoimansab lie,n mendop mlfoaiycnetaodlr i+ianl l5yw0 wh miictshh, elevation (degrees)−11000 two task conditions (single and dual) and ten experi- mental sessions. The single task condition required −20 only the performance of the head-slaved tracking task, −30 while the dual task required participants to perform the −40 tracking task and the monitoring task concurrently. −40 −30 −20 −10 0 10 20 30 40 azimuth (degrees) The order of the time delay condition was randomized Fig. 1 Example of target motion pattern for the head-slaved tracking across participants, while the order of the task condi- task. tions was fixed within each session (i.e., blocks of sin- Head position and orientation were measured by an gle-task trials preceded the dual-task trials). Ascension Bird tracker. The Bird consists of a DC magnetic-field transmitter and a receiver that was Apparatus and Procedure mounted atop the HMD. The Bird provides six de- grees-of-freedom tracking at 120 Hz while minimizing Each experimental session included 20 5-min interference caused by nearby metallic objects. All head-slaved tracking trials. The first 10 trials served as phases of the head-slaved tracking task and data col- a baseline condition for head-slaved tracking per- lection were governed by a 200 MHz personal com- formance and did not require the participant to perform puter. Target and head position data were collected at Each 5-min trial was preceded by a 5 s target acquisi- 60 Hz for each 5-min trial. tion period to ensure that participants had acquired the The nominal time delay in the head-slaved tracking target at the onset of the trial. Participants completed system was determined to be 46 ms. The imposed time the Simulator Sickness Questionnaire (SSQ; Kennedy delay conditions consisted of either three or six frames et al., 1993) at the completion of each 5-min trial and of delay, i.e., 50 or 100 ms, in addition to the nominal received a 10-min rest period after the completion of time delay. five experimental trials. Performance efficiency on the The secondary monitoring task consisted of the two tasks, as well as the associated workload data, have systems monitoring task from the Multi-Attribute Task been presented elsewhere (Nelson, Bolia, Russell, Battery (MATB; Comstock & Arnegard, 1992). In Morley, & Roe, 2000). short, the task comprised a set of four gauges with moving pointers. Under non-signal conditions, the Results moving pointers oscillated around the center tickmark by no more than one mark from the center tickmark on Responses to the SSQ were scored according the each of the gauges. A critical signal consisted of any procedures outlined in Kennedy et al. (1993). Scored of the four pointers moving more than one mark from in this way, the SSQ yields an overall index of simula- the center of the gauge in either direction. Participants tor sickness, referred to as Total Severity and three were instructed to inspect the gauges for critical signals subscales of simulator sickness – Nausea, Oculomotor and to make the appropriate keyboard response as soon Disturbance, and Disorientation. as one was detected. Critical signals not detected within 10 s were scored as missed signals; conversely, Total Severity Scores responses to non-signals were scored as errors of com- mission. The MATB monitoring task also included a Mean Total Severity scores were submitted to a pair of system status displays positioned above the four 3 (time delay) × 2 (task complexity) × 10 (experimental gauges. The normal or non-signal condition for these trials) repeated measures analysis of variance displays were the presence of a green light on the left (ANOVA), revealing a significant main effect for ex- display and a black fill on the right display. Critical perimental trials, F(9,54) = 2.30, p < .05, and a signifi- signals consisted of the left display shifting from green cant task complexity × experimental trials interaction, to black, or the right display shifting from black to red. F(9,54) = 2.26, p < .05. All other sources of variance Again, participants were instructed to inspect the sys- lacked significance; however, the time delay × experi- tem status display for critical signals and to respond as mental trials interaction approached significance, soon they detected a change in system status. During F(18,108) = 1.65, p > .05. The task complexity × ex- each of the 5-min experimental trials 12 critical signals perimental trials interaction, which is illustrated in Fig. were presented – two critical signals for each of the 3, can be explained by noting that in the single task four gauges and two critical signals for each of the condition Total Severity scores increased across trials, system status displays. but remained invariant across trials under the dual task condition. This impression was confirmed by post hoc test of the simple main effects of trials within the single and dual task conditions, F (9,54) = 3.18, p < .025, and F (9,54) = 0.67, p > .025, respectively. 15 )Q Task Complexity S Fig. 2 Participant performing head-slaved tracking task and secon- (recoS 12 DSiunagll eT aTsaksk dary visual monitoring task. S 9 y itre Upon arrival, participants were presented with an v overview of the experimental procedure, received in- l eaS 6 structions, and donned the HMD. Proper fit and view- to T ing quality in the HMD were achieved by making ad- n 3 a justments to its inter-pupillary distance controls, verti- e M cal, tilt, and axial helmet angles, chin strap, variable- 0 thickness foam pads, and inflatable air-bladder. Par- 0 1 2 3 4 5 6 7 8 9 10 11 ticipants completed 20 5-min head-slaved tracking tri- Experimental Trials als per experimental session – ten trials with and with- Fig. 3 Mean Total Severity scores as a function of trials under the out the additional visual monitoring task (see Fig. 2). single and dual task conditions. Inspection of Fig. 4, which illustrates the time de- lay × experimental trials interaction, indicates that the 10 nominal time delay condition was associated with the g highest Total Severity scores, and that these ratings n generally increased across experimental trials. Further ita 8 R examination of Fig. 4 reveals that the nominal + 100 ms e time delay condition resulted in the second highest rat- la cs 6 ings of overall simulator sickness, but remained rela- b u tively stable across experimental trials. It can also be S observed in Fig. 4 that the nominal + 50 ms time delay a e 4 condition was initially associated with the lowest sick- su a ness ratings, but eventually increased to a level that was N commensurate with the nominal + 100 ms delay condi- n 2 a tion. A post hoc analysis of these data indicated that e M the interaction can be explained by noting that sickness ratings varied across experimental trials in the nominal 0 0 1 2 3 4 5 6 7 8 9 10 11 + 50 ms time delay condition, F(9,54) = 2.67, p < .017, Experimental Trials but remained unchanged in the other time delay condi- tions (p > .017, n.s.). Fig. 5 Mean Nausea subscale ratings across experimental trials. 15 ) Time Delay Q S s 20 S 12 Nominal gn (e Nominal + 50 ms ita 18 ro Nominal + 100 ms R ySc 9 cn e 16 itre ab 14 veS 6 trsu 12 lt oa i roD 10 n T 3 mto 8 a e o 6 M lu 0 cc 4 O 0 1 2 3 4 5 6 7 8 9 10 11 n 2 Experimental Trials a e M 0 Fig. 4 Mean Total Severity scores as a function trials under the three 0 1 2 3 4 5 6 7 8 9 10 11 time delay conditions. Experimental Trials Simulator Sickness Subscale Ratings Fig. 6 Mean Oculomotor Disturbance ratings across experimental The Nausea, Oculomotor Disturbance, and Disori- trials. entation data were analyzed by three, 3 (time delay) × 2 Lastly, Fig. 7 illustrates the task complexity × ex- (task complexity) × 10 (experimental trials) repeated perimental trials interaction for the Disorientation measures analyses of variance, which indicated main scores. Perusal of the figure indicates that overall lev- effects for the experimental trials factor for both the els of Disorientation were very low throughout most of Nausea (Fig. 5) and Oculomotor Disturbance (Fig. 6) the experimental trials, with the exception of trial 9 in subscales, F(9,54) = 2.02, p < .05, and F(9,54) = 3.80, the single task condition. Post hoc analyses of these p < .05, respectively, and a task complexity × experi- data, however, failed to reveal the source of the inter- mental trials interaction for the Disorientation subscale, action. F(9,54) = 2.10, p < .05. All other sources of variance Conclusion lacked statistical significance. It can be observed in Fig. 5 that Nausea ratings increased across trials 1-5 The present study represents an initial effort to and remained relatively constant through trials 6-10. evaluate the occurrence of simulator sickness resulting Conversely, inspection of Fig. 6 reveals that Oculomo- from the effects of time delay, task complexity, and tor Disturbance ratings increased steadily across all experimental trials. simulator sickness decreased across sessions for each of 6 the SSQ's three dimensions, as well as for the SSQ's s g Total Severity index. In addition, the largest drops in n ita sickness ratings occurred between the first and second R 5 Task Complexity session in the VE, a result that has also be reported by e la Single Task Kennedy et al (1993) for flight simulators. In contrast, cs 4 the results reported herein indicate that ratings of b Dual Task u simulator sickness increased across time. Clearly, fur- S n 3 ther research exploring the effects of time on task on o ita simulator sickness in see-through HMDs is warranted. tn What relevance do these findings bear on questions e 2 ir regarding the appropriateness of see-through HMDs for o s real-world applications? As noted in the introduction, iD 1 n one of the supposed advantages of see-through displays a is that they will enable users (e.g., medical students, e M 0 surgical residents, fighter pilots, etc.) to practice and 0 1 2 3 4 5 6 7 8 9 10 11 rehearse the highly precise perceptual-motor skills re- Experimental Trials quired by their mission. However, given our finding that reports of simulator sickness increased with time Fig. 7 Mean Disorientation ratings as a function of trials under the on task, it is unlikely that users would be compelled to single and dual task conditions. spend a sufficient amount of time training and rehears- ing mission scenarios. The time on task effect also time on task using a see-through HMD. Despite the calls into question the efficacy of see-through HMDs highly-controlled nature of this experiment, we believe for tasks that require operators to use augmented dis- that these results are pertinent to human factors re- plays for a prolonged amount of time. Such a result searchers and system designers who are considering the may be particularly important for human factors profes- incorporation of see-through HMDs in multi-task work sionals advocating the incorporation of see-through environments. displays in a variety of application domains. One of the most striking outcomes of this experi- ment was the effect of time on task on simulator sick- Acknowledgments ness ratings. Specifically, participants’ ratings of Total Severity, Nausea, and Oculomotor Disturbance varied The authors wish to acknowledge the technical directly with this factor. In addition, the time on task contributions of Michael Poole and Liem Lu, of Sy- factor was found to interact with task complexity, such tronics, Inc. This work was funded by the Air Force that Total Severity scores associated with the single Office of Scientific Research under the New World task condition increased across experimental trials, Vista Research Program. while those associated with the dual task condition re- mained elevated, but stable. While the profile of Total References Severity scores as a function of time on task and time delay was more complex, it too revealed a general trend Adam, E. C. (1994). Head-up displays versus helmet- for increased sickness ratings as time on task increased. mounted displays: "The issues." Proceedings of the Collectively, these results suggest that overall symp- International Society for Optical Engineering: toms of simulator sickness were mediated primarily by Cockpit Displays, Vol. 2219 (pp. 13-19). the time on task factor, and not by time delay or task Bellingham, WA: SPIE-The International Society complexity. This is a surprising outcome given that for Optical Engineering. time delay has been proposed as one of the main etio- Azuma, R., & Bishop, G. (1994). Improving static and logical factors in the occurrence of simulator sickness dynamic registration in an optical see-through in HMDs. This also lends credence to the notion that HMD. In Computer Graphics Proceedings, Annual time delay per se is not sufficient for the onset of Conference Series, 197-203, ACM SIGGRRAPH. simulator sickness, a position that has received empiri- Beal, C. & Sweetman, B. (1994). Helmet-mounted cal support from the research of Draper et al. (in re- displays: Are we jumping the gun? International view) and Nelson (1996). Defense Review, 9, 69-75. It is also interesting to view these results in light of Comstock, J. R., & Arnegard, R. J. (1992). The Multi- reports that symptoms of sickness generally diminish as Attribute Task Battery for human operator workload a function of time spent in flight simulators (Kennedy and strategic behavior research. (Tech. Memoran- et al., 1993) and virtual environments employing dum 104174). Hampton, VA: NASA Langley Re- HMDs (Regan, 1995; Regan & Price, 1994). For ex- search Center. ample, Regan et al. found that participants' ratings of Draper, M. H., Viirre, E. S., Furness, T. A., & Gawron, virtual reality. Aviation, Space, and Environmental V. J. (in review). The effects of virtual image scale Medicine, 65, 527-530. and system time delay on simulator sickness within Rheingold, H. (1991). Virtual reality. New York: head-coupled virtual environments. Manuscript Summit Books. submitted to Human Factors. So, R. H. Y., & Griffin, M. J. (1991). Effects of time Durlach, N. I., & Mavor, A. S. (Eds.). (1995). Virtual delays on head tracking performance and the bene- reality: Scientific and technological challenges. fits of lag compensation by image deflection. AIAA Washington, DC: National Academy Press. Flight Simulation Technologies Conference, AIAA Furness, T. A. (1986). The super cockpit and its hu- Paper 91-2926, 124-130. New York: American man factors challenges. Proceedings of the Human Institute for Aeronautics and Astronautics. Factors Society 30th Annual Meeting. (pp. 48-52). Stinnett, T. A. (1989). Human factors in the super Santa Monica, CA: Human Factors Society. cockpit. In R. Jensen (Ed.), Aviation psychology. Hettinger, L. J., & Riccio, G. E. (1992). Visually in- (pp. 1-37). Brookfield, VE: Gower Publishing. duced motion sickness in virtual environments. Wells, M., & Griffin, M. (1987). A review and investi- Presence, 1, 306-310. gation of aiming and tracking performance with Kennedy, R. S., Allgood, G. O., & Lilienthal, M. G. head-mounted sights. IEEE Transactions on Sys- (1989). Simulator sickness on the increase. AIAA tems, Man, and Cybernetics, vol SMC-17, 210-221. Flight Simulation Technologies Conference, AIAA Worch, P. R. (1996). United States Air Force Scien- Paper 89-3269, 62-67. New York: American In- tific Advisory Board Study on UAVs: Technolo- stitute for Aeronautics and Astronautics. gies and Combat Operations. (Report No. SAF/PA Kennedy, R. S., Fowlkes, J. E., & Lilienthal, M. G. 96-1204). Washington D. C.: General Printing Of- (1993). Postural and performance changes follow- fice. ing exposures to flight simulators. Aviation, Space, and Environmental Medicine, October, 912-920. Kennedy, R. S., Lane, N. E., Lilienthal, M. G., Ber- baum, K. S., & Hettinger, L. J. (1992). Profile analysis of simulator sickness symptoms: Applica- tion to virtual environment systems. Presence, 1, 295-301. Kennedy, R. S., Lanham, D. S., Drexler, J. M., Massey, C. J., & Lilienthal, M. G. (December, 1995). Cy- bersickness in several flight simulators and VR de- vices: A comparison of incidences, symptoms pro- files, measurement techniques and suggestions for research. Paper presented at the Framework for Immersive Virtual Environments and Presence: Teleoperators and Virtual Environments Confer- ence (FIVE ’95), University of London. Kennedy, R. S., & Stanney, K. M. (1996). Postural instability induced by virtual reality exposure: De- velopment of a certification protocol. International Journal of Human-Computer Interaction, 8, 25-47. Nelson, W. T. (1996). Perceptual adaptation in a time-delayed virtual environment. Unpublished doctoral dissertation, University of Cincinnati, Cin- cinnati, OH. Nelson, W. T., Bolia, R. S., Russell, C. A., Morley, R. M., & Roe, M. M. (2000). Head-slaved tracking in a see-through HMD: The effects of a secondary visual monitoring task on performance and work- load. To be presented at the Human Factors and Ergonomics Society 2000 Annual Meeting, San Di- ego, CA. Regan, E. C. (1995). Some evidence of adaptation to immersion in virtual reality. Displays, 16, 135-139. Regan, E. C., & Price, K. R. (1994). The frequency of occurrence and severity of side-effects of immersion