Skill acquisition and task performance in teleoperation
using monoscopic and stereoscopic video remote viewing
Skill Acquisition and Task Performance in Teleoperation
using Monoscopic and Stereoscopic Video Remote Viewing
Industrial Engineering
University of Toronto
4 Taddle Creek Road
Toronto, Ontario
Canada, M5S 1A4
drascic@ie.utoronto.ca
Proceedings of the Human Factors Society 35th Annual Meeting,
1367-1371, San Francisco, September 1991.
(Winner of the Alphonse Chapanis Award for Best Student Paper)
(c) Copyright 1991.
Abstract
There are many tasks hazardous to human life which can be accomplished
remotely through telerobotic control. Robot technology has advanced to
the stage where teleoperated manipulators are versatile and effective
enough to be used successfully in a wide variety of circumstances. As
telerobotic systems become more sophisticated, it is important to
ensure that the human-machine interface is adequate for the task. One
very important type of feedback information that is missing from
standard telerobotic control stations is the immediate and compelling
binocular coding of depth, which is thwarted through the use of
standard monoscopic ( "2D") video systems, making the operator
dependent on other less salient visual depth cues. This is
unfortunate, since most telemanipulation tasks require operators to
have a good sense of the relative locations of objects in the remote
world.
To that end, a practical Stereoscopic Video (SV) system was developed
that is compatible with standard video display and recording
equipment. An experiment was conducted to examine the potential
benefits of SV for teleoperation. The results showed that SV can aid
teleoperation by reducing task execution times, reducing error rates,
and reducing the time needed for training.
Introduction
There are many tasks hazardous to human life which can be accomplished
remotely using telerobotic manipulators. The technology has advanced
to the stage where telerobots are are versatile and effective enough to
be used in a wide variety of circumstance. These teleoperators (also
known as remote manipulators and teleoperated vehicles) are in use
around the world, in the nuclear industry, police and military forces,
in space, and undersea. The majority of these telerobots use single or
multi-camera monoscopic video (MV) displays for feedback. (Meieran,
1988)
As Don Norman says in The Psychology of Everyday Things,
"Nothing succeeds like a good display." (Norman, 1988) Presenting
necessary information in a natural form facilitates all human-machine
interactions. Unfortunately, MV displays act as a filter, removing all
stereoscopic depth cues from a scene while retaining some of the
monoscopic cues. Stereoscopic cues are a very important source of
information about the spatial layout of the remote scene, and thus much
of the "knowledge in the environment" is rendered inaccessible. While
it is often possible to accomplish a teleoperation task using such a
display, it is usually difficult, and takes considerable training to
acquire sufficient skill. Robinson (1984) reports that standard
two-dimensional video systems with their restricted amount of visual
information resulted in Bomb Squad personnel being reluctant to use the
telemanipulation vehicle.
The loss of stereoscopic depth information means that there are
frequently times when the spatial locations of objects in a static
image are ambiguous. While motion parallax or multiple views can
sometimes resolve such ambiguities, operating conditions may render
this option unfeasible.
Dependence on monoscopic cues.
MV displays filter out all stereoscopic cues, and several monoscopic
cues as well, such as texture gradient. Operators therefore have a
greater dependence on certain monoscopic depth cues than most
stereoscopically-able persons are accustomed to.
It has been shown that when using a modified direct view, such as with
a prism arrangement to artificially exaggerate eye separation, or
magnifying lenses, monocular cues need to be learned, or
recalibrated, a process which takes time (McGovern, 1987).
Furthermore, it is known that binocular depth cues play a fundamental
role in the calibration of the monocular depth cues, and that binocular
disparity is perceived more quickly than any other visual cue (Clapp,
1986, Clapp, 1987).
Even after having learned how to interpret monocular cues for a
considerable time, it remains a fairly weak depth cue, and is easily
dominated by other cues such as perspective and occlusion. (Wickens,
1990)
On the other hand, as Baker notes: "Because the accommodation and
convergence differs in stereoscopy and the physical world, the ability
to see binocular depth on a CRT must be learned. On first occasion,
many people adapt in a few seconds, while others may take several
minutes to see the image comfortably." (Baker, 1987) This short time
period, however, is likely considerably shorter than the time needed to
master interpretation of monocular cues.
These facts imply that it will take a novice longer to become
proficient in teleoperation using a monoscopic display than a
stereoscopic display. Since the view from the video monitor is very
different from a direct view of the real world with respect to the
relationship between the monocular depth cues and the binocular depth
cues, it is reasonable to expect that without constant practice, the
temporary voluntary recalibration (or learning) of the depth cues of
the monoscopic display will fade.
For degraded or complex displays, the monoscopic cues may not prove
sufficient; other more radical methods of obtaining depth information
may be required: "At present [using MV displays] robots used in the
nuclear industry and elsewhere have to make contact with their
surroundings for the operator to know exactly where they are. Aided
only by a standard two-dimensional TV picture, the operator has to bash
the robot arm around inside the reactor until the right position is
somehow established. `This can cause damage both to the robot arm and
the surroundings.'" (Macilwain, 1989)
Benefits of stereoscopic displays
The expected benefits of using stereoscopic displays include a faster
and more accurate perception of the spatial layout of the remote scene,
visual noise filtering, enhanced effective image quality, enhanced
slope and depression detection, wider field of view, enhanced object
recognition and image interpretation, increased user satisfaction, and
fewer casual errors (critical movements are generally done with such
care that there is very little room for improvement). (Drascic, 1991,
Milgram et al, 1989, Merritt, 1988)
The literature shows that using stereoscopic video (SV) can greatly
improve teleoperation performance and user satisfaction, particularly
for tasks "which involve ballistic movement, recognition of unfamiliar
scenes, analysis of three dimensionally complex scenes and the accurate
placement of manipulators or tools within such scenes." (Dumbreck et
al, 1987) SV can in fact make possible tasks that are otherwise
impossible (Merritt, 1984).
Further, it has been quite clearly shown that as image quality is
degraded or scene complexity increased, the advantages of SV are
intensified. (Pepper, 1986)
As stated above, a certain amount of training is necessary to learn how
to use the monoscopic depth cues on a MV display. This suggests
another benefit of SV displays not greatly discussed in the literature,
which is that they do not require the same degree of training, and
should be easier for novices to use, possibly reducing the training and
practice time needed for skilled teleoperation. While the parameters
affecting the utility of SV displays have been investigated for a
variety of stereoscopic display formats (e.g. Pepper, 1983 Spain,
1984), most studies have examined the behaviour of well-trained
operators. Very little work has been done to investigate the behaviour
of relative novices to remote manipulation tasks, particularly with
regards to the type of video system used. Of the studies that have
been carried out, the findings have not been consistent. For instance,
whereas one study found that for a simple target positioning task, the
advantages of using a SV system were not as great for novices as for
experienced operators, implying that the relative benefits increase
with experience (Pepper & Hightower, 1984), another study, using a
peg-in-the-hole task, found, in contrast, that the relative benefits of
SV decreased with experience (Smith et al, 1979).
In order to investigate this question further, the author and his
colleagues conducted two experiments in the Department of Industrial
Engineering at the University of Toronto. The first (reported in
Drascic, 1991) investigated the behaviour of novices to teleoperation,
and found that for a very simple task with little need for depth cues
of any kind, using SV displays provided an initial advantage over MV
displays, an advantage which faded as subjects became more accomplished
at using the MV displays for the highly repetitive task. The second
experiment, reported here, examined skill acquisition for skilled
operators under a variety of different conditions.
Stereoscopic Video (In)Dependence
Merritt states: "Certain visual tasks are trivially easy with 3D
vision, but extremely difficult with only 2D vision. In some cases,
the monocular cues do not carry the required depth information; in
other cases the monocular cues normally present are degraded by poor
visibility or image quality." (Merritt, 1984) Extending this idea,
teleoperation tasks can be seen as existing in a spectrum between the
two extremes of not requiring any depth information (i.e.
"SV-independent"), and being impossible without stereoscopic displays
(i.e. "SV-dependent").
Driving a telerobot along a clear path would fall towards the
SV-independent end of the spectrum: the information required in order
to accomplish the task can easily be obtained without SV. The
experiment alluded to above involved such a task, and found that the
benefits of SV faded as subjects became more accustomed to using a MV
display. (Drascic, 1991) It is reasonable to expect that all tasks
near the SV-independent end of the spectrum will show similar results
(e.g. Pepper et al, 1981).
On the other hand, at the SV-dependent end of the spectrum, there will
likely exist tasks where the initial advantage of SV might even
increase with experience, as skill with SV improves, while skill
with MV is so handicapped that little improvement can occur.
Between these two extremes will exist the bulk of teleoperator tasks.
Since most telerobots in use today are equipped with MV system(s)
(Meieran, 1988), the only tasks possible have been those near the
SV-independent end of the spectrum. As SV becomes more commonly used,
the variety of telerobotic tasks will increase dramatically.
Purpose of the Experiment
The experiment was designed to examine the issue of skill acquisition
of a highly repeatable task. Furthermore, in order to see if the
benefits of SV were dependent on the difficulty and the SV-dependence
of the task, the highly repeatable task was designed so as to have
well-calibrated different difficulty levels, examining a broad
cross-section of SV-dependence levels.
Experimental Hypotheses
A previous study investigating the effects of degraded image quality on
a telerobotic task using MV and SV found that, for the easiest
condition, those subjects using MV showed considerable learning, while
those using SV displayed no clear learning effect. The authors argue
that since the task was designed to have very strong monoscopic cues
(i.e. be near the SV-independent end of the spectrum), it is not
unexpected that SV shows no improvement; the main improvement for the
MV condition is likely through acquiring skill in using the MV display,
skill they already had for SV displays. (Pepper et al, 1981)
Based on these results and the discussion above, the hypotheses posed
for this experiment were:
Experimental Hypotheses
1. Subjects using SV will show an initial performance advantage over
those using MV.
2. The performance difference between MV and SV will decrease as the
subjects become more experienced, more so for the low difficulty
SV-independent task conditions than the high difficulty SV-dependent
task conditions.
Method
Experimental Design
Design Goals. This experiment consisted of two parts, A and B, and
took place across two days. On Day 1 of the experiment, the subjects
performed both parts of the experiment using one of type of video
system. On Day 2 (with at least one day intervening) the subjects
repeated the experiment using the other video system.
Part A of the experiment was designed to look at the acquisition of
highly task-specific skills for a very repetitive situation, as a
function of experience (trial number), the video system used, and the
difficulty of the task. Part B of the experiment was designed to look
at the differences in performance of tasks in a non-repetitive
situation, as a function of the video system used and of the difficulty
of the task.
In Part A, the factors being examined were (1) task learning, by having
the subjects repeat the same task 16 times consecutively; (2) video
system, either SV or MV; and (3) task difficulty and SV-dependence,
using four different target sizes (8, 16, 32, and 64 cm). A
full-factorial design was used, so that each subject performed 16 * 2 *
4 error-free trials in Part A. Trials with errors were
discarded (as explained below), so this meant that subjects with a high
error rate performed more runs in total than those subjects with low
error rates. Although this may influence the results somewhat, since
some subjects therefore had more experience and practice than the
others, it was felt that the additional experience from making errors
would not contribute a great deal to the performance of the task, since
errors were so disruptive (see below).
In Part B, the factors being examined were (1) video system, either SV
or MV; and (2) task difficulty, with one of the four target sizes.
Each of the 2 * 4 conditions was repeated 8 times, so each subject
performed 2 * 4 * 8 error-free trials in Part B of the experiment.
Again, those subjects with high error rates performed more trials than
those with low error rates.
Final Design.
Considering the role of experience and transfer effects, Part A had the
following factors:
Order (2 levels: MF ("Mono First") and SF): between groups
Video (2 levels: MV ("Mono Video") and SV): within groups
Difficulty (4 levels): within groups
Learning (16 levels): within groups
Part B had the following factors:
Order (2 levels: MF and SF): between groups
Video (2 levels: MV and SV): within groups
Difficulty (4 levels): within groups
Replication (8 repetitions in a randomised order)
Experimental Procedure.
All subjects participated in the experiment on two non-consecutive
days. On Day 1 they performed the entire training and set of trials
using one video system, and on Day 2 they repeated the training and the
set of trials using the other video system. Four of the subjects (MF
"mono first" Group) began with the MV system, the other four (SF
"stereo first" Group) with the SV system.
The task was derived from an X-raying procedure used by the Canadian
Department of National Defence Explosive Ordnance Disposal team in
examining suspected parcels. (See Drascic, 1991) The task was
simplified, and a Fitts' Law approach was used to control and calibrate
the level of difficulty and SV-dependence. The telerobot used for this
experiment was the Remote Mobile Investigation unit (RMI), manufactured
by Pedsco (Canada) Ltd. The RMI is a mobile platform with a three
degree of freedom manipulator arm and an 80 m tether. Using various
tools and attachments, the RMI can be used to remotely X-ray, disable,
detonate, or transport a suspected explosive device, without risk to
human life. The RMI is typically equipped with a single MV system.
The one used for this experiment was modified to have a switchable
MV-SV system.
The experiment task consisted of driving the RMI a distance of 3 m
forward, and lowering a mock X-Ray photographic plate between two
"bombs", set a certain distance apart. The X-ray photographic plate
was simulated by using a pointer suspended from the end of the RMI's
forearm, that could swing freely. The two "bombs" were flat black
briefcases. The operators began each condition with the forearm of the
RMI pointed upward, so that the target was not visible on the monitor
screen. The operators had to lower the arm until the target was
visible, drive forward until the hanging pointer was between the two
briefcases, and lower the forearm until a buzzer on the pointer
sounded, indicating the end of the trial. (See Figure 1)
Figure 1: Fitts' Law Task
The subjects were told that the "bombs" were touch sensitive, so that
touching either suitcase would be counted as an error. This was done
because pilot studies revealed that accidently touching either suitcase
would very often require the operator to make complex recovery
actions. And more importantly, the design of the control panel for the
RMI was ergonomically poor: the toggle switch controlling the movement
of the forearm of the RMI was upside down with respect to stereotypes
of control-response compatibility of most of the subjects. When most
subjects in the pilot studies and in this experiment committed an error
by lowering the arm onto one of the suitcases rather than between then,
their quick, instinctive response was to pull the toggle switch towards
themselves, in the hopes of raising the forearm of the robot.
Unfortunately, this control action causes the RMI's forearm to lower
even further, resulting in a disruption of the experimental apparatus.
Because of the massive interference in task execution times caused by
errors, they had to be discounted from the data. Every time a subject
made an error, an additional run was added, so that the total number of
successful runs was constant for all subjects. The number of errors
was recorded for each set of trials.
According to Fitts' Law, movement time is linearly related to the Index
of Difficulty (ID), measured in bits, where ID = log2 (2 *
movement distance / target width). Using this formulation, the
different target sizes (bomb separations) can be converted into Index
of Difficulty bits. Given a movement distance of 3.0 m, and target
widths of 0.64, 0.32, 0.16, and 0.08 m, the corresponding IDs are 3.2,
4.2, 5.2, and 6.2 bits.
Training.
At the start of the experiment, each subject received a standard set of
instructions, describing the experiment and their task. It was
emphasized that speed was important, but that every error meant they
would have to perform another trial and should to be avoided.
This was followed by a familiarisation period with the controls of the
RMI. The RMI was then placed in the standard starting position, and
the briefcases were set to the training separation of 24 cm. Using
direct view the subjects practiced the experimental task until they
were able to pass a skill level criterion by performing four
consecutive error-free trials in under six seconds. They then repeated
the above familiarisation period and training procedure using the
remote view.
The subjects were eight university students, three female, five male.
All were volunteers, and were paid $5 per hour for their
participation.
Stereoscopic Video System.
An alternating field SV display with active liquid-crystal shuttering
spectacles was used for this study. (Milgram & van der Horst,
1986) The cameras for this experiment were set approximately 10 cm
apart, with their longitudinal axes converging on the tip of the
hanging pointer, approximately 95 cm away. This setting permitted good
depth resolution without causing undue eyestrain.
Measures Recorded.
The two measures recorded were the trial execution time, and the trial
success. It is important to stress that in this experiment both the
trial execution time and the number of faulty trials are
relevant measures of performance. This is a task subject to a speed
accuracy trade-off. In order to avoid making the subjects very
cautious, the importance of speed was emphasised. Because it is
impossible to know exactly where on the speed-accuracy trade-off curve
the subjects are at any given moment, it is important to examine both
sets of results.
Observations and Discussion
Training.
All subjects were trained to the same performance-based level of
performance using both direct view and remote view, as discussed
above. Based on Hypothesis 1 above, it was expected that those
subjects using SV on their first day would finish their training
faster. Figure 2 shows the mean number of trials taken to pass the
training criterion of four consecutive error-free trials in less than
six seconds.
Figure 2: Training
An analysis of variance of these data finds that the three-way
interaction of Order, Day and Video is significant at the 10% level
(F(1,6)=4.402, p=.081). As we can see from Figure 2, the only
difference between the two groups of subjects occurs on Day 1, when the
subjects are first learning how to use the telerobotic device
remotely. Here we see that those subjects using SV take an average of
16 trials to pass the training criteria, while those using MV take an
average of 28. This is clear support for Hypothesis 1, and suggests
that SV can be of great aid for novices to telerobotics.
Part A: Consecutive Presentation.
In order to clearly observe any trends, the data are grouped into four
sets of four trials and presented below in Figures 3 and 5.
Furthermore, since performance must be measured in both speed and
accuracy, the corresponding error rates are shown in Figures 4 and 6.
The error rate in these figures refers to the number of trials that
were rejected during the time it took to gather the four successful
ones averaged together in the corresponding Figure (3 or 5).
Examining first the results of Day 1 of the experiment, we find that
there is considerable support for Hypothesis 1, that subjects using SV
have a considerable performance advantage over those using MV at
first.
Looking at the easiest condition (ID = 3.2) in Figures 3 and 4, we see
that the trial times for SV are considerably shorter than those for
MV. Furthermore, there is a considerable downward trend in the MV
times. The error rates for both MV and SV are relatively low and
approximately equal for both MV and SV. The advantage of SV is
decreasing throughout the set of 16 trials.
At ID = 4.2, the next harder condition, we see that the trial times for
both MV and SV are slower than for the previous condition, as
expected. MV shows a decreasing trial time throughout the 16 trials,
with a consistent error rate. SV shows a slightly decreasing trial
time, but has an increasing error rate, suggesting that the subjects
are exploring the speed-accuracy trade-off more than exhibiting signs
of learning. Again, the advantage of SV appears to decrease throughout
the set of 16 trials.
At ID = 5.2, there is an indication of learning with MV (the second
group of 4 trials have a decreased error rate while trial time remains
the same; the third and fourth groups show fewer errors still and
slower trial times which might suggest a simple speed-accuracy
trade-off, or might indicate continued improvement in performance),
while those using SV show no particular learning trend (decreasing
error rates are matched by slower trial times, suggesting a
speed-accuracy trade-off). The advantage of SV does not appear to
decrease throughout this set of 16 trials, unlike the previous two
conditions.
At ID = 6.2, both MV and SV show a strong learning trend in the error.
The MV time drops for the second grouping of four trials, and then
rises for the two groups, while the error rate continues to drop,
suggesting some exploration of the speed-accuracy trade-off. SV shows
a small decrease in trial time with a large drop in error rate at
first, followed by constant times and an increasing error rate.
This could be an indication of fatigue, or simply a statistical
artifact. Again, the SV advantage does not appear to decrease
throughout this set of 16 trials.
Figure 3: Part A Day 1 Trends
Figure 4: Part A Day 1: Number of Errors versus Trial Number and Index of
Difficulty with Std Error Bars
In summary, then, we find that for the easier conditions there is
considerable improvement in performance (as measured by both time and
error rate) for the MV condition, with very little change in the SV
condition. At the higher levels of difficulty the learning trends are
less obvious for both MV and SV, and there appears to be more active
exploration of the speed-accuracy trade-off.
Furthermore, the advantage of SV decreases as the subjects become more
experienced, especially for the easier conditions. This is consistent
with hypothesis 2.
Now consider the results of Day 2 of the experiment. The subjects are
very much more experienced with the task on this day, albeit with the
other type of video system.
These results are considerably different from those of Day 1. A
cursory examination of the trial times shows no particular advantage of
SV except for the most difficult condition. A glance at the error
rates, however, shows that there is indeed a difference in performance
between MV and SV.
At ID = 3.2, there is little difference in the times, but those using
SV had a lower error rate for the first few trials. Those using MV
showed some learning in both time and error rate, so that there was
very little difference in performance by the end of the set of 16
trials.
At ID = 4.2, those using MV showed little change. Those using SV
appear to get worse briefly, then slightly better. The small SV
advantage at the beginning vanishes by the end of the 16 trials.
At ID = 5.2, those using MV anomalously perform considerably better
during the first set of four trials than the rest, getting
considerably worse and then little better. This suggests that the
initial set of trials were unusually good, a statistical
anomaly. Those using SV get consistently better, trading off time for
errors a little. Although the performance of MV is better than SV at
first, this situation is quickly reversed to the expected situation
with SV performance being consistently better than MV. As on Day 1,
the SV advantage does not appear to decrease with experience.
At ID = 6.2, those using MV show consistent learning, predominantly in
error rates. Those using SV performed better at first than in the rest
of the set of trials, similar to the MV performance at ID = 5.2. Here
the SV advantage does decrease, but does not vanish, with experience.
Figure 5: Part A Day 2 Trends
Figure 6: Part A Day 2, Number of Errors versus Trial Number and Index of
Difficulty, with Std Error Bars
Analyses of Variance (not shown) of these results confirm the
statistical significance of these observations. There is indeed a
consistent benefit from SV on Day 1 at all difficulty levels being seen
in reduced trial times, but there is no indication of a similar benefit
on Day 2 in the trial times. Further analysis of variance reveals that
there is no consistent effect on there error rates due to video, but
there is a consistent difference due to the task difficulty: the harder
the task, the more errors are made. There is also a significant trend,
or "learning", reported for Day 1. For Day 2, the same effect is
significant at the 10%. Given the behaviour of MV at the ID = 5.2
level and SV at the ID = 6.2 level, this lowering of significance is
not surprising.
In general, then, the results of Part A of the experiment strongly
support Hypotheses 1 and 2.
Part B: Random Presentation
Part B, where the separation between the two targets was varied
randomly with each trial, was designed to examine a more steady-state
behaviour than that in Part A, with subjects who were well-skilled at
the task, but had not necessarily memorised the particular motor
movements necessary for that particular task, and did not know the
precise conditions of the next trial. Although 8 trials for each
separation and video combination were performed, no learning trend was
expect or subsequently found in the data.
Figure 7 shows the results of the trials times of Part B of the
experiment as a function of the Index of Difficulty, while Figure 8
shows the same for the error rate results.
Figure 7: Part B (Random Presentation) Trial Times
Figure 8: Part B Percent Errors
If we first examine the results of Day 1 of the experiment, we find the
very unexpected results that the SV advantage for trial times actually
decreases as the difficulty of the task increases.
However, if we examine the corresponding plots of errors, we see that
the difference between MV and SV is considerably different at the most
difficult task level. Considering that performance is a combination of
both trial time and error rate, we see that there is indeed a
consistent benefit due to SV. It appears that the subjects considered
speed to be far more important than accuracy. At that level of
difficulty, the low speed could only be obtained at the expense of a
high error rate.
Considering Day 2 of the experiment, we find that performance for both
MV and SV is much improved, thanks to the large amount of experience
received on Day 1. There is no significant difference between the MV
and SV conditions with regards to trial times, although those using MV
appear to be somewhat faster, at the expense of higher error rates.
The only significant difference between the MV and SV performance is at
the highest level of difficulty. Thus we find that the SV advantage
decreases quickly for SV-independent tasks, and is persistent for a
longer period for SV-dependent tasks.
Analysis of variance on these results (not shown) confirm the
observations made from the graphs above.
Conclusion
In the work described herein, various theoretical and practical aspects
of using SV rather than MV for teleoperation were considered. It was
hypothesised that SV is easier to to learn than MV, and that the
benefits in performance due to SV depended on how dependent the task
was on stereoscopic depth cues. For situations where these cues are
relatively unimportant, it was suggested that the benefits of SV would
be temporary, and last only as long as it took the operators to learn
how use the monoscopic depth cues of a MV display. For situations
where the stereoscopic depth cues are important, it was
suggested that the benefits of SV would last longer.
The experimental results support these hypotheses, and demonstrate that
the benefits of SV, even after a great deal of practice, will still be
apparent for difficult, SV-dependent tasks, long after the benefits
have faded for easier, SV-independent tasks.
The implications for telerobotics are obvious. Given the nature of
most telerobotics applications, operators have only a very few chances
to accomplish the task correctly. The performance benefits of SV, even
though they may fade with practice for highly repeatable tasks, should
be very strongly evident in these single-attempt situations.
Furthermore, given that operators can learn to use a SV display much
more quickly than a MV display, operators should require less initial
training and less constant practice in order to maintain their skills
at a suitable level.
Acknowledgements
The work described here was carried out under contract
W7711-7-7009/01-SE with Supply and Services Canada for the Defence and
Civil Institute of Environmental Medicine. The author would like to
thank his colleagues Dr. Paul Milgram of the University of Toronto, and
Dr. Julius Grodski of DCIEM for their assistance in this work.
References
Baker, J, "Generating images for a time-multiplexed stereoscopc
computer graphics system", SPIE Vol 761 True 3D Imaging
Techniques and Display Technologies, 44-52, 1987
Clapp, R E, "Stereoscopic Displays and the human dual visual
system", SPIE Vol 624 Advances in Display Technology VI,
41-52, 1986
Clapp, RE, "Stereoscopic Perception", SPIE Vol 761 True 3D
Imaging Techniques and Display Technologies, 79-87, 1987
Drascic, D
"An Investigation of Monoscopic and
Stereoscopic Video for Teleoperation ",
MASc Dissertation, University of Toronto, June 1991
Dumbreck, A A, C W Smith, S P Murphy "The Development &
Evaluation of a Stereoscpoic Television System for Use in the Nuclear
Industry", Int'l Workshop on Nuclear Robotic Technologies and
Applications, University of Lancaster, June/July 1987
Macilwain, C "Remote control robots seen through 3D spectacles"
, THE ENGINEER, 35, 8 June 1989
McGovern, D E "Current developments needs in the control of
teleoperated vehicles", SANDIA National Laboratories Report
SAND87-0646 UC-15, Albuquerque, New Mexico, August 1987
Meieran, H B "Robotics and teleoperator-controlled devices",
Health Physics, v 55. n 2, 215-222, Aug 1988
Merritt, J O, "Visual tasks requiring 3-D stereoscopic
displays", SPIE Vol 462 Optics in Entertainment II, 56-59,
1984
Merritt, J O, "Often-overlooked advantages of 3-D displays",
SPIE V. 902 Three-Dimensional Imaging and Remote Sensing
Imaging, 46-47, 1988
Milgram, P, D Drascic, J Grodski "Stereoscopic Video + Superimposed
Computer Stereographics: Applications in Teleoperation", Proc.
Second Canadian Workshop on Military Robotic Applications,
Kingston, Ontario, Aug 1989.
Milgram, P, R van der Horst "Alternating-field stereoscopic
displays using light-scattering liquid crystal spectacles",
Displays: Technology & Applications, v 7, n 2, 67-72, April
1986
Norman, D A, The Psychology of Everyday Things, Basic Books Inc,
New York, 1988
Pepper, R L, "Human Factors in Remote Vehicle Control", 30th
Annual Meeting of the Human Factors Society, Dayton, Ohio, Sep-Oct
1986
Pepper, R L, J D Hightower "Research Issues in Teleoperator
Systems", Proceedings of the HFS 28th Annual Meetings, pp 803-807,
1984
Pepper, R L, David C Smith, Robert E Cole "Stereo TV improves
operator performance under degraded visibility conditions",
Optical Engineering, v 20, n 4, 579-585, July/Aug 1981
Pepper, R L, "Research issues involved in applying stereoscopic
television to remotely operated vehicles", SPIE Proceedings, v 402,
170-173, 1983
Robinson, M "Remote control vehicle guidance using stereoscopic
displays", Proceedings of the HFS 28th Annual Meeting, 809,
1984
Smith, D C, R E Cole, J O Merritt, R L Pepper "Remote Operator
Performance Comparing Monoa nd Stereo TV Displays: the Effects of
Visibility, Learning, and Task Factors", NOSC Technical Report
No. 380, Feb 1979
Spain, E H, A Psychophysical Investigation of the Perception of
Depth with Stereoscopic Television Displays, PhD Dissertation, U of
Hawaii, 1984
Wickens, C D "Three-dimensional stereoscopic display
implementation: Guidelines derived from human visual
capabilities", SPIE Vol 1256 Stereoscopic Displays and
Applications, 2-10, 1990