Poster Session 2 Summaries
Please
note: Some
summaries have been edited for space and clarity. The conference proceedings
will contain complete abstracts and papers.
(1)
Traffic
Maneuver Problems and Crashes of Young Drivers Adam Kirk, Nikiforos Stamatiadis (University of Kentucky ¾ USA)
While over the past decades the population of
younger drivers has been decreasing, their crash rates have increased. Past research has associated their
higher crash rates to societal influences and youthful behavior. The objective of this research is to
identify the specific driving maneuvers whose unsuccessful undertaking results
in specific types of crashes involving these drivers. Four types of crashes were identified as the most prominent
for young drivers including crashes at intersections, rear end, crashes
resulting from passing maneuvers and single vehicle crashes. The analysis was performed examining
the Kentucky crash database for the 1994-1996 period using the quasi-induced
exposure method. The results showed
that for all crashes there is a general trend of decreasing involvement with
increasing age, which indicates that their inexperience is the largest single
contributor to their increased crash rates. Of significance is the fact that for all crashes a dramatic
decrease of involvement after the first year of driving between the years of 16
and 17 is observed. This may be
indicative of a steep learning curve in the first years of driving regarding
the ability to control a vehicle.
Therefore, very little can be made to improve this phenomenon. Increasing the level of awareness among
young drivers about these issues and their likely crash involvement seems to be
the only viable approach. However,
preliminary efforts from the graduated license show that some of these trends
seem to be reduced indicating a possible impact on the crash rates of young
drivers.
(2)
Driver
License Renewal Issues and Concerns Nikiforos Stamatiadis (University of Kentucky ¾ USA)
Periodic renewal of driver licenses is an integral
part of the licensing procedures for most states, including Kentucky. Renewal of driver licenses is usually
required every four years, and many states conduct vision tests before granting
renewal. A few states require
additional testing, while several states, including Kentucky, have no vision or
any other examinations at renewal.
Past research has shown a relationship between crashes and driving
records. Thus, it was considered
important to develop mechanisms to identify potential problem drivers and to
systematically review current practices regarding license renewal and
retesting. Additional concerns
included the increasing percentage of elderly drivers and the deterioration of
their vision due to aging. There is a universal agreement that vision plays a
significant role in driving performance, that there are age-related visual
changes, and that drivers over age 75 have proportionally higher crash rates
than younger drivers. However,
there is no established standard for vision-screening policies. Age-based road
tests are not considered to be a practical means for identifying drivers with
deficiencies, and they would unnecessarily burden the license renewal
process. However, using road tests
as an additional means of evaluating select individuals, such as those failing
vision tests or those referred by a physician or family member, could significantly
improve the identification of deficient drivers. The work completed here
indicates that safety gains might be achieved by implementing additional
procedures for older drivers. Such
procedures might include requiring drivers over 75 to renew their licenses
every 2 years, and using vision screening tests that include a set of medical
questions for older drivers. Allowing renewal examiners the discretion to
require road testing as deemed necessary is also recommended. Despite our considerable knowledge
about the physiological changes of older persons and the impact of these
changes on driving, further research is needed. Such research should seek to design and evaluate license
renewal programs that would provide older persons with a fair assessment of
their driving abilities.
(3)
Detection
of Collision Events by Older and Younger Drivers George John Andersen, Asad Saidpour, AnnJudel
Enriquez (University of California, Riverside ¾ USA)
Recently (Andersen et al., 2000; 1998) we found
that older drivers performed more poorly than younger drivers in a situation
that required them to detect an impending collision during braking. In the present study, we examined
whether older drivers performed poorly compared to younger drivers in detecting
a collision with a moving object. Twenty-two older and younger drivers were
presented with computer-generated scenes of a roadway in a driving simulator.
Located in the scene was a single object that moved independently of the
vehicle motion, and that was or was not on a collision path with the
vehicle. Overall older drivers
were less sensitive to detect a collision than younger drivers, with
performance worse for long as compared to short time-to-contact (TTC)
conditions.
(4)
The
Role of Simulation in a Staged Learning Model for Novice Driver Situational
Awareness Training
Loren Staplin
(TransAnalytics, LLC ¾ USA), James C. Dowdell (SafeDrive Technologies ¾ USA)
This paper theorizes that an optimal strategy for
training novice drivers to acquire situational awareness skills will rely on a
hierarchical approach consistent with traditional models of cognitive
development. The success of
applying such models hinges upon information presentation techniques that can
maximize depth of processing, and hence comprehension and retention, at a
specific stage of learning. Our
general discussion argues that the appropriate use of simulations is uniquely
suited to meet this need.
(5)
Driving
Tests: Reliability and the Relationship Between Test Errors and Accidents
Chris Baughan, Barry Sexton
(TRL Limited (Transport Research Laboratory) ¾ UK)
In the British practical driving test, serious or
dangerous faults are those judged to involve potential or actual danger, and a
single such fault results in test failure. As part of a wider project to review the driving test, TRL
conducted a study of test-retest reliability. Test
and retest outcomes differed for a substantial proportion of candidates. The
paper argues that inconsistent performance on the part of the candidate is
likely to explain much of this apparent unreliability. Minor faults are recorded during the test but few candidates make the
16 of them necessary to fail. Self-reported accidents during the first
six months of post-test driving were analysed together with driving test records for approximately
30,000 drivers. Statistical modelling suggested that people who pass the
driving test having made large numbers of minor faults may be intrinsically
less safe as drivers, but that they also tend to drive less overall, and less
often at night. This reduces (and
for men removes) the association between test faults and the actual number of
accidents reported.
(6)
Feasibility
of Evaluating Design Ideas for Reducing Vehicular Entrapment at Railroad
Crossings Using a Laboratory Experiment One-Jang Jeng, Tirthankar Sengupta, Satya Vallepalli (New Jersey Institute of
Technology ¾ USA)
The number of accidents at railroad crossings is
particularly high at places where streets run parallel to the railroad tracks.
Existing grade crossings were investigated for potential problems and studied
for design solutions. The present study reports progress of the first phase of
a NJ DOT-sponsored project. A laboratory experiment was conducted to evaluate
various design ideas before they are implemented in a second-phase field study.
The laboratory study used images taken from actual scenes of railroad crossings
in New Jersey instead of the graphical drawings commonly used in driving
simulations. Possible design ideas were edited using image-processing software.
Design ideas were then saved in different layers in order to generate design
combinations that could be superimposed on the background images to create
virtual railroad crossing scenes. Nighttime images were also made possible by
retouching the digital daytime images. Preliminary results of the in-lab
experiment are presented. Lessons learned from the current project indicate
that the use of actual images with superimposed design ideas is a
cost-effective approach to evaluating and redesigning display layouts.
(7)
Modeling
Driver Cognition
Delphine Delorme
(University of California at Berkeley ¾ USA)
Modeling driver cognition is a challenging but
necessary endeavor in the effort to develop systems able to support drivers in
their decisions and actions. This paper focuses on the design and
implementation of the PADRIC (PATH DRIVer Cognitive) model, and on the module
in charge of reproducing part of the perceptive processing of the model. The
PADRIC model is integrated within a micro-simulation tool (SmartAHS) to support
the development and assessment of driver assistance systems. For example, it
has been used to generate simulations of visual distraction for highway-driving
resulting in hard braking maneuvers. Future research will seek to include
additional cognitive processes in order to increase the range of behaviors that
can be simulated and to introduce new knowledge databases in order to apply the
model to other traffic situations than highways.
(8)
Acceleration
Behavior of Drivers in a Platoon Ghulam H. Bham, Rahim F. Benekohal (University of Illinois at Urbana-Champaign ¾ USA), Outstanding
Student Paper Award Winner
A new dual-regime acceleration model was developed
to represent the acceleration behavior of drivers in a platoon of vehicles. Two
sets of field data collected by aerial photographic techniques were used to
assess the validity of the proposed and existing acceleration models. A single
regime acceleration model failed to present the acceleration behavior of
drivers. The field data indicated that at around 13 m/sec the acceleration rate
drops. Thus, two different acceleration rates, higher acceleration rate at
lower speeds and lower acceleration rate at higher speeds, were used to provide
the best fit to the data. This provided realistic acceleration behavior of
drivers in a platoon. The field data sets were collected about 10 years apart.
The improvements in acceleration capability of a platoon of vehicles from two
different time periods were determined. Improvements in performance of vehicles
were quantified using the above mentioned field data. The method of
quantification can also be used to predict and model the performance of
vehicles currently in use. Inversely, current vehicles can be downgraded to
represent vehicles of past years and thus make use of already collected data.
Important uses of the dual regime model are in modeling the traffic flow
behavior and designing roadway elements that depend on acceleration behavior of
drivers.
(9)
Evaluating
the Presence of In-Vehicle Devices on Driver Performance: Methodological Issues Christian Jerome, H.C. Neil Ganey, Patrick
Commarford, Brian Oakley, Mustapha Mouloua, Peter A. Hancock (University of
Central Florida ¾ USA), Outstanding
Student Paper Award Winner
A central concern of Intelligent Transportation
Systems (ITS) is the effect of in-vehicle devices (e.g. cell phones, navigation
systems, radios) on driver performance and safety. As diverse and innovative
technologies are designed and implemented for in-vehicle use, questions
regarding the presence and use of these devices assume progressively greater
importance. Concern for the safety of advanced driver training and require us
to develop and validate reliable and effective procedures for assessing such
effects. This work examines a number of candidate procedures, in particular the
evaluation of cognitive workload as a strategy by which such goals might be
achieved.
(10) Driver and Driving Assessment
Issues Associated with the Application of a Secondary Task Technique: A Case
Study Michael P.
Manser, Jacqueline Jenkins
(Texas A & M University ¾ USA)
This paper presents the results of an examination
of driver assessment techniques applied to a case study investigating the
relationship between conversation intensity while using a cell phone and driver
performance. A secondary task
technique was applied to study the influence of the intensity of conversation
on the degradation of driving performance attributed to using a cellular
telephone. Forty participants
drove through simulated driving environments and engaged in cellular telephone
conversations with the experimenter.
After driving through each environment participants provided a rating of
the workload experienced during the cellular telephone conversation during the
drive. Accelerator input, speed, steering input, and lane position measurements
were recorded while participants drove through particular sections of the
simulated driving environments.
Results of the study indicated the variation of steering input increased
when the cellular telephone was used.
Males generally drove faster, and females responded more quickly to a
situation requiring a sudden braking or steering maneuver. Overall, the drivers perceived the
workload to be greater when using a cellular telephone. However, the use of the
case study to examine the driver assessment and secondary task techniques
indicated there were several positive and negative attributes which need to be
considered in future research where such techniques are to be applied. For example, an analysis of the performance
measures revealed that changes in the horizontal alignment of the roadway
created lasting perturbations in the data. In particular, increased variations in steering input and
lane position were observed for a considerable distance and time after a
participant exited a curved section of roadway. These perturbations subsequently interfered with the
application of the secondary task.
(11) Evaluation of a Low-Cost,
PC-Based Driving Simulator to Assess Persons with Cognitive Impairments Due to
Brain Injury Jerry Wachtel (The Veridian Group, Inc. ¾ USA), William K. Durfee (University of Minnesota ¾ USA), Theodore J. Rosenthal (Systems Technology, Inc. ¾ USA), Elin Schold-Davis (Sister Kenny Institute ¾ USA), Erica B. Stern (University of Minnesota ¾ USA)
Brain injury due to accident or stroke frequently
results in cognitive impairment, reducing an individual’s ability to
judge driving situations accurately. And such individuals may lose the
metacognitive skills necessary to be aware of their own limitations. Typical
on-road evaluations conducted by rehabilitation professionals are generally
unable to assess the candidate’s responses to real-world driving
challenges because they are conducted under non-demanding conditions. Indeed,
individuals with mild cognitive deficits may perform adequately on such tests
but unsafely when driving challenges increase. In a project funded by the
National Center for Medical Rehabilitation Research, National Institutes of
Health, we sought to determine whether a low-cost, PC-based driving simulator
could provide clinicians with information useful to their efforts to assess the
safe ability to drive of individuals with cognitive impairments. We developed
two driving scenarios and pilot-tested them on ten subjects—five with
moderate impairments and five controls. Both the “simple” and
“complex” scenarios matched the essential route characteristics of
an existing on-road evaluation, but the “complex” scenario
incorporated common but demanding driving challenges, including: cross-traffic
failing to stop at a STOP sign; pedestrians crossing the driver’s path;
vehicles suddenly entering traffic from the shoulder; opposing traffic
appearing from behind slower moving vehicles; slow lead vehicles causing
passing decisions; and oncoming traffic forcing gap acceptance decisions for
left turns. Results from both discrete events and continuous performance data
were encouraging. We evaluated four discrete events: run-off-road; crashes;
failure to stop at STOP signs; and failure to execute directed turns. Whereas
each of the brain-injured subjects committed at least one of these errors, none
of the control subjects committed any. We collected continuous data for speed
maintenance, speed deviation, and lane position, in both the simple and complex
segments of the scenario, for two tangent and four horizontal curve sections.
Generally, the non-impaired subjects improved their performance during the
course of the 19-mile scenario, despite the fact that the complex events all
occurred during the second half of the drive, whereas the impaired
subjects’ performance degraded during the complex segment of the run. The
results will lead to an enhancement of simulator capabilities and a
comprehensive clinical trial at multiple sites.
(12) Driver Advocate TM Tool Chip Wood, Robert Leivian, Noel Massey, Jack Bieker, John
Summers (Motorola Labs ¾ USA)
Using scenario driven research, a Driver AdvocateTM
(DA) system has been designed to advise the driver about potentially unsafe
situations based on information from environmental sensors. DA is an intelligent dynamic system
that monitors, senses, prioritizes, personalizes, and sends alerts to the
driver appropriate to the moment.
This has the potential to sharply decrease driver distraction and
inattention. To support the realization of DA, a DA Tool (DAT) has been
developed to coordinate with a KQ (previously Hyperion) virtual driving
simulator and allow the merging of the simulated driving performance, the
environmental sensors, and the intelligent use of audio, visual, and tactile
feedback to alert the driver to potential danger and unsafe driving behavior.
DAT monitors the traffic, lane following, forward and side clearances, vehicle
condition, cockpit distractions, Infotainment use, and the driver affective
behavior. The DAT is designed to
be highly configurable, flexible, and user friendly to facilitate creative
freedom in designing usability and human factors experiments and rapid
prototyping.
(13) A HMD-Based Virtual Reality
Driving Simulator
Ronald R. Mourant (Northeastern
University ¾ USA),
Maria T. Schultheis (Kessler Medical Rehabilitation Research and Education
Corporation ¾ USA)
Recent advances in optics, HMD design, 3D graphics
chips, and processes for personal computers have combined to make HMD-based virtual
reality driving simulators available at low cost. A HMD with a resolution of 1,024 by 768 with a FOV of 50o
diagonally is now available for about $20,000. A graphics processor that can
render large databases at fast frame rates costs only $400. Personal computers can now support
multiple processors that run over 1 Gigahertz. We discuss visual concerns with
a HMD, choosing a HMD for a driving simulator, HMDs compared with fixed
displays, consequences of improved frame rates, autonomous vehicles, and the
use of a HMD based driving simulator for studying drivers who have cognitive
impairments.
(14) Meta-Analysis of Crash Risk
Factors Among Older Drivers: Application to a Model Program of Driver Screening Karlene Ball, Virginia Wadley, Jerri
Edwards, David Ball (University of Alabama at Birmingham ¾ USA), Daniel Roenker (Western Kentucky University ¾ USA)
In the absence of disease or impairment, there is
no empirical evidence that subtle, age-related changes in sensory or cognitive
function affect older drivers’ abilities to safely operate a motor
vehicle. However, impairments that
do affect driving occur with a higher prevalence in the older population. This
paper describes a meta-analysis of risk factors for automobile crash. Risks associated with compromises in
visual acuity, contrast sensitivity, visual fields, useful field of view, and
mental status are examined. Risks associated with medical conditions,
medications, and physical limitations also are discussed. Results demonstrate
that visual acuity and contrast sensitivity are only weakly associated with
crash risk, while cognitive variables and mental status measures are moderately
associated with crash risk. A
specific measure of The Useful Field of View, UFOVÒ,
is strongly associated with crash risk.
These results suggest that driver evaluations aimed at detecting unsafe
drivers can be significantly improved.
Specifically, evaluations can be improved by including a broader
assessment of visual function, an assessment of cognitive function including the
UFOVÒ,
an assessment of mental status, and an assessment of physical status. An evaluation incorporating these
components might facilitate driver qualification or referral for appropriate
intervention. A driver screening evaluation program incorporating these
components is currently underway.
With over 2,000 older adult participants, results from this model
program indicate that cognitive function, rather than visual or physical
function, is most strongly associated with mobility outcomes. Thus, interventions to maintain or
improve cognitive function may also help to sustain mobility
(15) Effects of Speed of Visual
Processing Training upon Non-Visual Attention in "At-Risk" Older
Drivers Nicole Skaar, Matthew Rizzo, Kirk Bateman,
Steven Anderson (University of Iowa ¾ USA)
Purpose: Reduction in a measure known as the
Useful Field of View (UFOV) is a risk factor for car crash involvement in older
drivers (Owsley et al, 1991; Ball et al, 1993). This measure depends on aspects
of visual attention (divided attention [DA], selective attention [SA]) and
speed of processing (Owsley et al, 1991; Ball et al., 1993). UFOV scores can be
improved through speed of processing training (Ball, Beard, et al., 1988a,b),
and this improvement may transfer to enhanced driving performance (Roenker et
al, submitted). This preliminary analysis addresses the hypothesis that
training of visual speed of processing can improve performance in
attention-demanding tasks that are processed outside the visual domain. Methods:
Forty participants were enrolled in an
ongoing study of at-risk older drivers based upon reduction in UFOV scores
(measured using the Visual Attention Analyzer, 3000). Twenty-two participants
ages 66-87 (mean age = 74.4; mean MMSE = 28.4; 65.4% male) were randomly assigned
to a speed of visual processing training group and 18 participants ages 67-91
(mean age = 75.1; mean MMSE = 28.7; 52.6% male) were randomly assigned to a
control group (who trained to use the Internet). Each group participated in 10
one-hour training sessions. All 40 participants also performed on a version of
the Paced Auditory Serial Addition Task (Gronwall, 1977), which was
administered before and after training. Results: After training, the
speed of processing training group showed improvement in mean scores on all
UFOV subtests (DA 66.18ms, SA 244.68ms [with identification of a foveal target]
and 115.27ms [for a same/different foveal discrimination]; P<0.001 Wilcoxon
signed rank, all cases). The controls showed improvement only on the selective attention
subtest (70ms p=0.049). On the PASAT, the speed of processing training group
showed improvement in mean scores on the 2.4 second PASAT (8.6%, p=0.01).
Results improved at the 2.0 second interval task, but this difference was not
significant (4.4%, P=0.11) - possibly due to a floor effect at this difficult
speed. The control group did not improve their PASAT scores significantly at
either PASAT speed (3.0%, P=0.51 at 2.4 second and 3.8%; P=0.22 at 2.0 second
intervals). Conclusion: Training of speed of visual processing improved
performance on visual attention tasks as expected. We also found preliminary
evidence of crossmodal transfer of these training effects leading to
improvement on the PASAT, a task that commands both auditory attention and
working memory. These preliminary findings suggest that speed of visual
processing intervention may entrain attention resources at supramodal levels
outside the visual modality. Improvement of attention in tasks outside the
visual domain after training of visual speed of processing may be relevant to
performance in tasks during driving such as multitasking, using a cell phone,
and engaging in conversation with a passenger.
(16) A Computational Model of Driver
Decision Making at an Intersection Controlled by a Traffic Light Terry Stanard, Robert J.B. Hutton (Klein
Associates, Inc. ¾ USA), Walter Warwick, Stacey McIlwaine, Patricia L. McDermott (Micro
Analysis and Design ¾ USA)
An important challenge
associated with driving simulation development is the computational representation
of agent behaviors. This paper describes the development of a preliminary
autonomous agent behavior model (based on the Recognition-Primed Decision (RPD)
model, and Hintzman’s multiple-trace memory model) mimicking human
decision making in approaching an intersection controlled by a traffic light.
To populate the model, an initial Cognitive Task Analysis was conducted with
six drivers to learn the important cues, expectancies, goals, and courses of
action associated with traffic light approach. The agent model learns to
associate environmental cues (such as traffic light color) with expectancies of
upcoming events (like light color change) and appropriate courses of action
(such as decelerating). At present, the model is currently being evaluated for
its successful representation of the Recognition-Primed Decision Making
process.
(17) Investigating Drivers' Traffic Knowledge
in Jordan Wa'el
Awad, Mohammad Rasoul
S. Al-kharabsheh (Al-Balqa - Applied University ¾ Jordan)
Jordan’s fatality rate per registered vehicle
is approximately 7.5 times larger than that of the United States (157.2 per
100,000 in Jordan vs 21.0 per 100,000 in the U.S.). This project addresses the
traffic safety problem in Jordan by evaluating driver’s knowledge of
existing traffic laws and regulations. An experiment was conducted in which 55
subjects with current driver’s licenses were administered a test composed
of 25 questions selected from actual Jordanian driver’s license exams.
Statistical analyses were then conducted on the results. It was found that a
shocking 96.4% of the drivers in this study failed to pass the simulated
written driver’s license exam, with professional drivers scoring worse
than non-professional drivers. Based on the findings, recommendations are made
regarding Jordanian public policy governing driver’s licensing, including
more frequent retesting of drivers, a higher standard of knowledge for traffic
rules, and a nationwide program to assess the relationship between driver
knowledge, driver behavior, and crash and fatality rates.
(18) Socioeconomic Characteristics of
Speeding Behavior
Kyungwoo Kang (Hanyang
University ¾ Korea)
Many studies on drivers’ speeding behavior have
been reported in the last decade. Most of the previous studies, however, have concentrated
on the relationship between drivers' speeding behavior and road/vehicle
characteristics, without considering other important factors such as personal
characteristics and drivers' perception of the speed limit. This paper analyzes
Korean drivers' speeding behavior by taking into account such factors as trip characteristics
in addition to personal, vehicular, and attitudinal factors. Speeding behavior
is measured by a categorical measure over the speed limit, and an ordered
probit model is used to econometrically estimate the speeding behavior
equation. Results indicated that i) male drivers with higher income tend to
drive faster, and experienced drivers drive at higher speeds than others ii)
vehicles with more horsepower and vehicles with safety features go slower than
vehicles with less safety features iii) trip distance and frequent use of the
road are important factors for speed selection behavior, and iv) perceived
speed limit of the road and expectation of being caught for speeding are
important factors for driving behavior.
(19) Driver Alertness Detection
Research Using Capacitive Sensor Array Philip W. Kithil (Advanced Safety Concepts, Inc. ¾ USA)
The research project compared and analyzed
physiological and performance data for 13 subjects driving a vehicle simulator.
Each subject drove the simulator for morning, afternoon, and late night
sessions. These sessions were intended to represent alertness conditions during
an “awake” baseline period and the secondary and primary circadian
sleep cycle periods. The sessions were approximately one hour, two hours, and
two or three hours in length, respectively. With one exception, the subjects
had experienced normal sleep the night before the test. Five men and eight
women participated, ranging in age from 25 to 59. Physiological data included:
real-time PERCLOS (percentage of slow-eye closure over one minute) using an
infrared-reflective camera; head position coordinates using an overhead
capacitive sensor array; and video of the right front of the subject’s
face. Performance data included: vehicle speed, lane departures, lane
deviation, and steering/turn signal data. The research manager maintained logs
of unusual circumstances such as departing the roadway, falling asleep at the
wheel, excessive speeding, etc. Head position data was analyzed and compared to
the videos. A multi-element algorithm was developed which captured patterns of
head motion found to be characteristic of drowsiness. The algorithm output was
compared to roadway departures noted in the research manager’s logs of
unusual events. The comparison showed a capability of advance detection of
about 87% of driver roadway departures with a false positive rate of about
15%.
(20) The Influence of Conversation,
Low-Dose Alcohol and Driving Experience on the Peripheral Vision System Peter Langer, M. Kopp, B. Holzner
(University of Innsbruck ¾ Austria), W. Magnet (Kuratrium fuer Verkehrssicherheit ¾ Austria)
This study investigated whether legal everyday
occurrences which take place while driving a motor vehicle and which require
some attention, such as talking to a passenger or being under the influence of
a low dose of alcohol (between 40 and 50 milliliters alcohol level), influence
the scope of the driver's visual field. Alcohol in low doses does not damage a person's
eyesight but reduces cognitive attention. Also investigated was whether the
effect is related to driving experience. Methods: The peripheral vision
reaction time of 60 persons was measured using the "Peripheral Vision
Test" by Schuhfried. The test subjects were divided into three groups
(n=20 each): Group 1 was asked to hold a conversation during the test;
Group 2 took the test under the influence of a low dose of alcohol
measured from the subject's breath using the "Alcotest 7410"
(Dräger Sicherheitstechnik, Germany); Group 3 served as the control.
Each group was divided into two subgroups, namely persons with average driving
experience (more than 50.000 km) and persons having less driving
experience (less than 5.000 km). All statistical analyses were performed
on SPSS 8.1 for Windows. Differences between groups were tested for
significance by means of analysis of variance and the non-parametric
Mann-Whitney U Test. Results: Significant differences in the average
reaction time were seen between the control and the conversation groups (0.76
vs. 1.20 seconds; p=0.01) and also between the control and the alcohol groups
(0.76 vs. 1.03 seconds; p=0.04). This difference is enhanced when we look at
the reaction times in the subgroups, divided into experienced and
less-experienced drivers. The differences between the control group with
experience and the conversation group with less experience is highly
significant (p=0.003), with the experimental group having faster reaction time,
as is the difference between the control group with experience and the alcohol
group with less experience (p=0.004). The two alcohol subgroups also differ
significantly from the control group with experience in terms of the average
number of wrong reactions (0.8 vs. 1.8 ; p=0.029 and 0.8 vs. 2.8; p= 0.002). Conclusions:
Holding a conversation with a passenger while driving a car reduces the
peripheral vision field. The same effect can be observed in persons under the
influence of a low dose of alcohol. The effect is enhanced when the person also
has limited driving experience. The difference between a driver holding a
conversation and an inebriated driver is that the inebriated driver not only
has a longer reaction time but also shows more wrong reactions.
(21) Human Factors in Highway-Rail
Crossing Accidents: The Influence of Driver Decision Style Mansour Rahimi, Najmedin Meshkati
(University of Southern California ¾ USA)
This paper explores the hypothesis that driver
decision-making style influences highway-rail crossing accidents. To investigate this, we have designed
an analysis of variance experiment with three independent variables:
“driver decision style,” “driver time pressure” and
“intersection complexity.”
To simulate the driving conditions, we identified and videotaped a number
of dangerous crossings in downtown Los Angeles. The tapes represented different environmental complexities
and time pressures a driver experiences while crossing an intersection. The
tapes were played back to the subject drivers. The subjects were classified according to their decision
styles. Dependent measures were
designed based on a driver’s decision to cross the intersection. This paper presents the conceptual
approach and the experimental design for this research.
(22) Fatigue Countermeasure Using Automatic
Real-Time Video Processing of Eye Characteristics Jeffrey B. Bishop (Future of
Technology and Health, LC ¾ USA), Isaac K. Evans (Evolutionary
Heuristics ¾ USA)
Fatigue is a large and growing problem for aviators
and motor vehicle drivers. A fatigue
countermeasure based on digital video processing of images of the subject’s face has been developed. Digital video data of drivers was
collected in experiments in a driving simulator at University of Iowa. Algorithms were developed to
automatically locate the head, eyes, and face features of the driver using wide
field-of-view images. The exposed
eye area is quantified using the final eye target location. Automatic processing of face
features in a vehicle environment is a difficult task due to the complexity of
the scene and the variable lighting conditions. The use of both static and dynamic processing in parallel
and the use of symmetry has lead to the development of innovative and useful
algorithms for automatic face location and feature detection. The prototype system was effective in
automatic feature location for all 13 subjects tested in driving simulator
studies. The appearance of visible
eye features is quantified and used to determine onset of potentially dangerous
fatigue conditions. An alarm event
is generated if the eyes remain closed longer than 1.5 seconds, or if sustained
reduction in area of exposed eye features is detected (eyelid droop). Other studies in the literature have
validated correlation of measures of pupil occlusion (such as PERCLOS) with
operator performance lapse. The
system is designed to work with an inexpensive digital video sensor mounted on
the dashboard of a vehicle and runs on standard computing hardware.
Copyright © 2001 University of Iowa Public Policy Center. All rights reserved.