Accident Analysis and Prevention 49 (2012) 7885

Contents lists available at SciVerse ScienceDirect

Accident Analysis and Prevention
j our nal homepage: www. el sevi er . com/ l ocat e/ aap
Are car daytime running lights detrimental to motorcycle conspicuity?
Viola Cavallo

, Maria Pinto
French Institute of Science and Technology for Transportation, Development and Networks, Laboratory of Driver Psychology, 25, Alle des Marronniers, 78000 Versailles, France
a r t i c l e i n f o
Article history:
Received 2 November 2010
Received in revised form28 July 2011
Accepted 12 September 2011
Keywords:
Motorcycle safety
Sensorial conspicuity
Attentional conspicuity
Daytime running lights (DRLs)
Vulnerable road users
a b s t r a c t
For a long time, motorcycles were the only vehicles with daytime running lights (DRLs), but this con-
spicuity advantage has been questioned due to the rapidly increasing introduction of DRLs on cars as well.
The present experiment was designed to assess effects of car DRLs on motorcycle perception in a situa-
tion that specically brought attentional conspicuity to bear. Photographs representing complex urban
trafc scenes were displayed briey (250 ms) to 24 participants who had to detect vulnerable road users
(motorcyclists, cyclists, pedestrians) appearing at different locations and distances. Car DRLs hampered
motorcycle perception compared to conditions where car lights were not on, especially when the motor-
cycle was at a greater distance from the observer and when it was located in the central part of the visual
scene. Car DRLs also hampered the perception of cyclists and pedestrians. Although the globally positive
safety effect of car DRLs is generally acknowledged, our study suggests that more attention should be paid
to motorcyclists and other vulnerable road users when introducing car DRLs. Several means of improving
motorcycle conspicuity in car DRL environments are discussed.
2011 Elsevier Ltd. All rights reserved.
1. Introduction
The question of visual conspicuity has long been considered
decisive for improving the safety of motorcyclists (e.g., Janoff and
Cassel, 1971; Ramsey and Brinkley, 1977; Woltman and Austin,
1974). There is currently a newsurge of interest inthis issue inmost
developed countries, which have been experiencing a substantial
increase in the death toll of motorcyclists for several years now,
to the point of becoming alarmingly high. The number of motor-
cyclists killed in Europe in 2008 was as high as 17% of the total
number of road fatalities (IRTAD, 2009), even though motorcycles
represent only 2% of road users (EC, 2009). In France, the propor-
tion of motorcyclists killed on the road rose in 10 years from9.2%
(in 1996) to 16.6% (in 2005), and the risk of a fatal accident per
kilometer travelled by motorcycle riders is 20 times higher than
for automobilists (Guyot, 2008). Similar trends have been noted in
Australia (Federal Ofce of Road Safety, 1997). In the U.S., the risk
of a fatal accident per kilometer travelled by motorcyclists as com-
pared to automobilists has even doubled, going from 18 times as
high for motorcyclists in 2000 (Tessmer, 2000) to 37 times as high
in 2007 (NHTSA, 2008).

Corresponding author at: Institut Franc ais des Sciences et Technologies des
Transports, de lAmnagement et des Rseaux (IFSTTAR), Laboratoire de Psychologie
de la Conduite (LPC), 25, Alle des Marronniers, 78000 Versailles, France.
Tel.: +33 01 30 84 39 44; fax: +33 01 30 84.
E-mail addresses: viola.cavallo@ifsttar.fr (V. Cavallo), maria.pinto@ifsttar.fr
(M. Pinto).
Accident studies (ACEM, 2009; Hurt et al., 1981) have shown
that collisions between motorcycles and other vehicles represent
the most frequent type of motorcycle accident (between 70% and
75%, according to ACEM and Hurt et al., respectively) and that in
two-thirds of such collisions the motorcycles had the right-of-way
(Hurt et al., 1981; Wulf et al., 1989). Most of these accidents occur
in urban areas (ACEM, 2009), in daylight, in clear weather, and
with a motorcycle bearing between 11 and 1 oclock relative to the
other vehicle (Brooks et al., 2005). The most typical automobile-
motorcycle accident happens when an automobile turns left (in
countries where one drives on the right) and crosses the path of
an oncoming motorcycle. In-depth accident studies point to the
high frequency of perception failures of the other vehicle drivers
(in 37%, 60%, and 70% of the cases according to ACEM, 2009; Van
Elslande and Jaffard, 2010; Hurt et al., 1981). The failures are most
often identied as resulting fromno detection or late detection of
the motorcycle. They are often referred to as look but failed to
see errors, where car drivers report that they looked in the right
directionbut did not see the approaching motorcycle (e.g., Crundall
et al., in press; Shahar et al., in press; Williams and Hoffmann,
1979). Some authors also mention another kind of failure: even
when the motorcycle was detected and identied, other vehicle
drivers may still misjudge its speed, distance, and time-to-arrival
(Horswill et al., 2005; Pai, 2011; Tsutsumi and Maruyama, 2008).
Regarding detectability, prior research has demonstrated that
motorcycles, by nature, are less conspicuous than cars (e.g.,
Hancock et al., 1990; Olson et al., 1981).
Conspicuity can be dened as the property of an object
to attract attention (Connors, 1975). Conspicuous objects are
0001-4575/$ see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aap.2011.09.013
V. Cavallo, M. Pinto / Accident Analysis and Prevention 49 (2012) 7885 79
easily and rapidly noticed (Engel, 1974). Two forms of conspicuity
are generally distinguished (Hancock et al., 1990). Sensorial con-
spicuity, also sometimes called salience (Itti and Koch, 2000), is
based on bottom-up processes that inuence selective attention,
and is determined by the combination of the physical properties
of the perceived object, such as its size, brightness, color, out-
line, and motion (Engel, 1977; Itti et al., 2003; Yee et al., 2001).
Background properties are essential insofar as they determine con-
trasts (of brightness, color, and motion) and visual noise (presence
of clutter and competing patterns). Cognitive conspicuity is based
on top-down attentional processes that depend on the observers
expectations in relation to her/his past experiences, intentions,
and motivations. A similar distinction was made by Hughes and
Cole (1984) between attention conspicuity and search conspicuity,
depending on whether an object is noticed by an observer whose
attention is not specically directed toward its possible occurrence
or whether the observer is expecting and deliberately searching for
it.
Concerning the detectability of motorcycles, it has been shown
that their sensorial conspicuity is reduced due to their small size,
their often dark colors, and their irregular contours. Their cogni-
tive conspicuity is poor because of low exposure frequencies, and
unusual behaviors such as high speeds (Brenac et al., 2006) and
unexpected locations (driving between lanes), which are inconsis-
tent with the other vehicle drivers expectations.
In the past, and still largely today, most of the research and
application efforts have been devoted to improving the sensorial
conspicuity of motorcycles. The main safety measure has been
the use of daytime running lights (DRLs) by motorcycles, which
became compulsory in the seventies in many countries. In France,
for example, this measure was introduced in 1975. A double effect
on motorcycle perception could be expected. At the same time as
motorcycle DRLs would increase the vehicles contrast with the
environment, they were likely to improve detectability by attract-
ing the attention of other road users. Furthermore, as long as
motorcycles were the only vehicles to turn on their lights during
the day, DRLs would provide a consistent feature, a sort of visual
signature that wouldmake it easier for other users to identify them.
Having DRLs on a motorcycle is considered to counteract its inher-
ently lower conspicuity, and the positive effect of this measure
in preventing motorcycle accidents has been proven in a number
of studies (Dahlstedt, 1986; Hendtlass, 1992; Muller, 1984; Olson
et al., 1981; Thomson, 1980; Williams and Hoffmann, 1979; Zador,
1985).
The conspicuity advantage of motorcycles as the only vehicles
with daytime lights has been questioned due to the growing prac-
tice among automobilists of turning on their lights during the day.
Car DRLs are now authorized in most countries and mandatory in
many others. In Europe, some formof car DRLs is mandatory in 17
E.U. countries, and from2011 onwards, ECE R48 legislation makes
dedicated daylight running lamps mandatory in the E.U. for new
cars and small delivery vans conforming to ECE R87. A number of
automobile manufacturers foresawthis event and began to market
vehicles with dedicated DRLs that go on automatically when the
vehicle is turned on and do not allowthe driver to choose whether
or not to use them. In this context, it seems important to determine
howand whether the use of daytime lights by automobilists alters
motorcycle conspicuity, and if so, to nd new ways of improving
motorcycle detectability.
Many researchers and road-safety specialists, as well as motor-
cyclist associations, argue that car DRLs may have a detrimental
impact on motorcycle safety (e.g., Brendicke et al., 1994; FEMA,
2006; Hrberg and Rumar, 1979; Knight et al., 2006). Indeed, car
DRLs arelikelytocompetewithmotorcycleDRLs bycreatingvisual
noise that hinders the perceptionof motorcycles. Anadverse effect
could also arise for cyclists and pedestrians: if cars with daytime
lights are more conspicuous, then unequipped users could become
less conspicuous andthus less noticeable (e.g., Brouwer et al., 2004;
Hrberg and Rumar, 1979).
The potential impact of car DRLs onroadaccidents has long been
a topic of debate among road-safety specialists and has given rise
to a large body of accident studies and some experimental research
(for a review, see for example Cairney and Styles, 2003).
While a large number of accident studies suggest that car DRLs
improve safety (e.g., Andersson and Nilsson, 1981; Elvik, 1996;
Koornstra et al., 1997; Tofemire and Whitehead, 1997), we agree
with Elvik (1996) in saying that the safety gain, generally estimated
at 510%, depends on the surrounding luminosity, making the gain
greater in Finland, for example, than in Israel. Several studies have
found little or no signicant improvement (Farmer and Williams,
2002; Theeuwes andRiemersma, 1995; Wang, 2008). Some authors
have pointedout the possible negative effects of car DRLs onunpro-
tected users (e.g., Theeuwes and Riemersma, 1995; Wang, 2008).
Others recommend using dedicated low-intensity car DRLs (rather
than standard passing beam headlights) to limit adverse effects
on motorcycles (e.g., Knight et al., 2006), and still others advise
implementing measures that increase the conspicuity of motor-
cycles (e.g., Rumar, 2003). However, most studies have not found
any detrimental effects of car DRLs (e.g., NHTSA, 2000; Paine, 2003;
Riemersma et al., 1987; Rumar, 2003; Schnebeck et al., 2005).
Surprisingly, only few experimental studies have investigated
the effect of car DRLs on the detection of vulnerable users, notably
motorcyclists. Regarding possible masking effects, Cobb (1992,
cited in Brouwer et al., 2004) studied various car DRL intensity
levels in real-world conditions. He found that DRLs improved the
perception of automobiles in cloudy weather without decreasing
that of motorcycles and bicycles, provided no high-intensity lamps
(greater than 600cd) were used. Brouwer et al. (2004), who used
slides of trafc scenes including one car and sometimes another
road user, even observed a slight detection advantage for the other
road user when car DRLs were lit. Brendicke et al. (1994) also used
slides of trafc scenes and observed a detrimental effect of car DRLs
on motorcycle detection, but the ecological validity of their nd-
ings is questionable insofar as the participants task was to count
the number of vehicles in the scene.
On the basis of previous studies it is therefore not clear whether
DRLs affect the detectability of motorcycles and other vulnerable
road users. All of these ndings suggest that these inconclusive
results might be due to methodological choices, and to how the
very notion of sensory conspicuity is operationalized. Most stud-
ies have used simplied experimental tasks and situations, such as
searching for a motorcycle in an impoverished environment and/or
with an unlimited or long exploration time. Such situations are not
representative of the attentional demands of real driving, however,
and are therefore much less likely to generate perception errors.
We contend that the sensorial conspicuity of motorcycles should
be assessed in complex environments, with a time-limited task
where the observer has to notice a motorcycle, not look for one.
In short, the situation must call upon selective attention, in such a
way that attentional conspicuityrather thansearchconspicuityis at
stake.
In the light of these considerations, the present investigation
looks at the effects of automobile DRLs on the perception of
motorcycles and other vulnerable road users. A complex urban
environment and low-luminosity conditions (overcast skies) were
used. The participants task was to detect various kinds of vulnera-
ble users motorcyclists, cyclists, and pedestrians in conditions
of rapid informationintake. The vulnerable road users were located
at different distances and eccentricities in the road environment so
as to limit the participants use of expectation-driven search strate-
gies. Target distance and eccentricity are known to be determining
factors in object detection (Engel, 1971; Jenkins and Cole, 1986;
80 V. Cavallo, M. Pinto / Accident Analysis and Prevention 49 (2012) 7885
Rog et al., 2009) insofar as near (big) and centrally located targets
are detected more easily than far (small) and off-centered ones.
We hypothesized that in the here studied conditions, the use
of daytime lights by automobilists not only creates visual noise
that decreases the detectability of motorcycles, but also hampers
the detection of other vulnerable road users. Car DRLs may even be
more harmful to unlit road users such as cyclists and pedestrians.
We also assumed that the effect of car DRLs is modulated by the
distance and eccentricity of the vulnerable road user that needs to
be detected.
2. Method
2.1. Participants
Twenty-four adults (4 women and 20 men) with a mean
age of 35.6 years (SD: 11.1) participated in the experiment.
They were all licensed drivers and had normal or corrected-to-
normal vision (at least 8/10 binocular acuity). All participants
underwent three visuo-attentional tests: UFOV Part 3 (Ball et al.,
1993) and Eye Movement and Covert Shift of Attention from
the Test for Attentional Performance (TAP, Zimmermann and
Fimm, 2002). All participants exhibited normal performance. The
participants were given a full explanation of the experimen-
tal procedures. Written informed consent was obtained before
participation.
2.2. Task
The participants were shown color photographs representing
road-trafc scenes. They had to detect the presence of vulner-
able road users, and if one was detected, determine whether it
was a motorcyclist, cyclist, or pedestrian. Three visual targets
were used to prevent the observers from looking specically for
motorcycles.
2.3. Experimental design
Arepeated measures design was employed, with all stimuli pre-
sented to all participants. The rst independent variable was DRL
(on vs. off) of all four-wheel vehicles in the scene (whereas motor-
cycles always had their front lights on). The second variable was
the targeted vulnerable road user, which could be a motorcyclist,
a cyclist, or a pedestrian. The third variable was the distance of
the vulnerable road user, who could be in a near or far location
with respect to the observers (the cameras) point of view. The
fourth variable was the eccentricity of the vulnerable road user,
who appeared in a central area of the photograph or off-centered.
Combining these variables led to a total of 24 experimental condi-
tions. Three different photographs ineachof the 24conditions were
presented in the experiment, making a total of 72 experimental
trials.
One hundred and eight distractor trials were added. Among
them, 18 trials contained a vulnerable road user who was located
at different distances from those used in the experimental trials
so as to vary the positions in the visual scene where a vulnerable
road user could appear. We therefore had a total of 90 trials con-
taining a target. The other 90 distractor trials (without a vulnerable
user) were added to balance the design and to have 50% of the
photographs contain a target. Distractor trials were not analyzed.
The 180 stimuli were presented in three blocks of 60 stimuli, each
block containing 24 experimental and 36 distractor trials. The pho-
tographs were presentedfor 250ms to simulate the amount of time
a driver might have when glancing in the direction of oncoming
trafc (Crundall et al., 2008).
2.4. Apparatus and stimuli
Photographs of real-world trafc scenes were used as stimuli.
They were presented on a 40

LCD at panel display (Samsung

SyncMaster PXn), which offered a high-quality visual rendering
(image contrast ratio of 1200:1, 1366768 resolution, brightness
500cd/m
2
, 16.7 million colors).
Digital photographs of urban trafc at intersections (mainly in
Paris) in overcast weather were taken using a Nikon D90 camera
with a 35-mmlens. Twenty-two intersections were photographed
at a height of 1.20m (average viewpoint of a driver), from differ-
ent vantage points and a large number of trafc congurations. The
idea was to obtain a wide variety of trafc scenes in order to reduce
participants expectations about the location of the visual targets
and prevent them from using a xed search strategy or heuristic.
At the same time, we wanted to control visual-scene complexity,
and to present the different targets in similar road environments
to ensure comparability. To do so, the photographs were catego-
rized according to their contextual features (kind of intersection,
number of lanes, viewing angle, line of cars parked or not along
the curb, etc.). These categories were validated in a pre-test. The
photographs were edited in Photoshop by a computer graphics
designer. Their luminosity and contrast levels were made uniform,
undesirable items (e.g., salient objects like billboards that might
interfere) were eliminated, and the various elements needed to
create the experimental conditions (e.g., vehicles and pedestri-
ans) were inserted. We were careful to keep the images realistic
so that the participants attention would not be attracted to any
incongruities.
2.5. Experimental stimuli
Thestimuli includeda number of cars (between3and6) stopped
at or approaching a red light, and a single vulnerable road user
(a motorcyclist, cyclist, or pedestrian), which was the target the
participants had to detect. The visual characteristics of the targets
were highly diverse, but their height in the image was controlled.
Vehicle headlights were edited into the pictures using Photoshop,
and included realistic halos and reections on the road surface
(Fig. 1); DRLs of cars and motorcycles were matched for luminosity.
If needed, the targets were resized so that their height on the view-
ing screen was 6cm(angular size 2.15

) or 4cm(angular size 1.43

)
for the near andfar conditions of the distance variable, respectively.
Target eccentricity was a function of the vulnerable users location
in the picture. Three locations were dened by drawing two imag-
inary vertical lines dividing the image into three equal parts. The
area between the two lines was called the central location, and the
areas to the left and right were called the off-centered locations.
2.6. Distractor stimuli
Two kinds of distractor stimuli were used. The rst type con-
tained vulnerable road users at distances that differed from the
experimental trials, and vehicles whose headlights were (all or
partly) onor off. The secondtype containedno targets, but depicted
a wide variety of trafc situations. They included trafc scenes
where some or all of the vehicles had their headlights on or off
(so that participants did not notice that DRL was an experimen-
tal variable). The vehicles could be at very different distances (in
the foreground or background) and in all kinds of positions with
respect to the observers viewpoint (turning right or left, etc.).
2.7. Procedure
All participants rst tooka visual acuitytest (Ergovision) andthe
TAP visuo-attentional test. Then they were seated 160cmfromthe
V. Cavallo, M. Pinto / Accident Analysis and Prevention 49 (2012) 7885 81
Fig. 1. Examples of photographs showing a motorcycle inthe near/central condition
and a DRL-off (upper) or DRL-on (lower) environment.
monitor. The images displayed on the screen subtended a visual
angle of approximately 32

18

. Next, the participants read the

instructions and performed a practice block of 8 trials to get famil-
iar with the procedure. The 180 stimuli were then presented in
3 blocks of 60 trials. The presentation order of the trials within
the blocks was random for all participants, and the order of the
blocks was counterbalanced across participants. Each trial started
with a xation point displayed in the center of the screen for
1500ms. Then a stimulus appeared for 250ms, followed by the
question Apart from the four-wheel vehicles, which other road
user did you see?. The answer choices were (1) a pedestrian, (2)
a motorcyclist, (3) a cyclist, (4) none of the above. The question
remained on the screen until the participants selected an answer
using the computer mouse. No feedback on the response accu-
racy was given. After an inter-stimulus interval of 500ms, the
next xation point was displayed. A short break was proposed
between the blocks. The experimental session ended with the
UFOV test. The whole experimental session lasted between 40 and
45min.
2.8. Data analysis
The participants responses were assigned to several response
categories. For the experimental trials (those trials that contained
a vulnerable road user), the number of correct detections, identi-
cation errors, and misses were determined: a correct detection
was counted when the right category of vulnerable road user was
recognized, an identication error was counted when a vulnerable
road user was detected but not correctly identied, and a miss was
counted when the vulnerable road user was not detected at all. For
the distractor trials (those trials that did not contain a vulnerable
road user), a distinction was made between false detections and
correct non-detections: false detections (false alarms) occurred
when a vulnerable road user was detected but no one was present,
Table 1
Mean percentage and standard error (SE) of correct detections by road user and DRL
condition.
DRL-Off DRL-On Total
Mean % SE Mean % SE Mean % SE
Motorcyclist 65.6 3.0 60.1 3.3 62.8 3.2
Cyclist 49.0 3.4 40.3 3.8 44.6 3.7
Pedestrian 53.5 2.7 47.9 3.0 50.7 2.9
Total 56.0 3.3 49.4 3.7 52.7 3.7
and correct non-detections occurred when a vulnerable road user
was not present and not detected.
Correct detections and identication errors were input into
within-subject ANOVAs. To this end, the scores on correct detec-
tions and identication errors were transformed to normality
(arcsin). AcompleteANOVAincludingall four variables couldnot be
conducted because data were not normally distributed in that case.
The effects of distance and eccentricity on correct detections were
analyzed by two separate one-way ANOVAs. Signicant effects
were further examined using Scheff tests. The effect size (
p
2
) was
alsocomputed. Supplementaryanalyses oncorrect detections were
conducted using McNemar tests to determine whether the car DRL
effect on vulnerable road-user detection depended on target dis-
tance and eccentricity. The analyses were carried out separately
for motorcyclists, cyclists and pedestrians. For each road user cat-
egory, data were partitioned into four sub-sets dened by near,
far, central and off-centered conditions. McNemar tests were
carried out for each of these 12 sub-sets of data, by comparing
paired detection scores (correct vs. incorrect detections) in DRL-
on and -off conditions. The -level was set at .05 for all statistical
analyses.
3. Results
3.1. Overall results
On the experimental trials, the correct-detection and
identication-error scores were 53% and 7%, respectively, whereas
the miss rate was 40%. On the distractor trials, the observed rates
were 5%for false detections and95%for correct non-detections. The
overall d

value of 1.73 indicates that the participants responses

clearly differed fromchance performance.
The ANOVA on correct detections with DRL and road user
as within-subject factors revealed main effects of car DRLs,
F(1, 23) =18.56, p<.0005,
p
2
=.45, and road user, F(2, 46) =32.09,
p<.0001,
p
2
=.58. No signicant interaction between car DRLs and
road user was observed. The signicant car-DRL effect therefore
indicates that all vulnerable road users were more often detected
when the vehicle head lights were off than when they were on (see
Table 1). Concerning the main effect of road user, post hoc analyses
revealed that motorcyclists were detected signicantly better than
were pedestrians and cyclists, and that pedestrians were detected
more often than cyclists (see Table 1).
The two one-way ANOVAs on correct detections with distance
andeccentricity as within-subject factors yieldedsignicant effects
of distance and eccentricity (p<.0001) with higher detection rates
when the vulnerable road user was located nearby (M=72%) rather
thanfar away(M=33%), andwhenlocatedinthe center of the visual
scene (M=69%) than when it was off-centered (M=37%).
A two-factor repeated measurement ANOVA on identication
errors yielded a main effect of car DRLs, F(1, 23) =22.07, p<.0001,

p
2
=.49, with more identication errors when car DRLs were on
(9%) than when they were off (5%).
82 V. Cavallo, M. Pinto / Accident Analysis and Prevention 49 (2012) 7885
0
20
40
60
80
100
Off-centered Central Far Near
Eccentricity Distance
D
e
t
e
c
t
i
o
n

r
a
t
e

(
%
)
Off
On
[*]
[*]
Fig. 2. Means and standard errors of the motorcyclist detection rate, as a function
of distance, eccentricity, and whether the car DRLs were on or off ([*]p<.1).
0
20
40
60
80
100
Off-centered Central Far Near
Eccentricity Distance
D
e
t
e
c
t
i
o
n

r
a
t
e

(
%
)
Off
On
[*]
*
**
Fig. 3. Means and standard errors of the cyclist detection rate, as a function of
distance, eccentricity, and whether the car DRLs were on or off ([*]p<.1, *p<.05,
**p<.01).
3.2. Effects of car DRLs according to vulnerable road user,
distance, and eccentricity
McNemar tests were conducted for each of the vulnerable road
users to determine whether the effect of car DRLs was mediated
by the distance and eccentricity of the visual target in the road
scene. Note that the presence of car-DRLs systematically had a neg-
ative impact onthe detectionof vulnerable roadusers, althoughthe
effect was not always signicant.
With regard to motorcycle detection, marginally signicant
effects of car DRLs were observed in the far condition (47% vs. 40%;
p=.08) and in the central condition (83% vs. 76%; p=06) (see Fig. 2).
Concerning the detection of cyclists, signicant car-DRL effects
were noted in the near condition (74% vs. 60%; p<.0025) and in the
off-centered condition (44% vs. 34%; p<.05). A marginally signi-
cant difference was also observed in the central condition (54% vs.
46%; p=08) (see Fig. 3).
As to pedestriandetection, there was a signicant car-DRL effect
in the central condition (81% vs. 72%; p<.05) (see Fig. 4).
4. Discussion
The present study revealed a detrimental effect of car DRLs on
the perceptionof vulnerable roadusers: motorcyclists, cyclists, and
pedestrians were less well detected when the headlights of cars in
their vicinity were on. Overall, the detection rate for vulnerable
road users decreased by seven percentage points when the cars
headlights were turned on. Note that the effect of car DRLs went in
the same direction in all distance and eccentricity conditions, that
is, detection performance always declined (although not always in
a statistically signicant way) whencar DRLs were on. Car DRLs also
increased the identication-error rate (the number of detected but
0
20
40
60
80
100
Off-centered Central Far Near
Eccentricity Distance
D
e
t
e
c
t
i
o
n

r
a
t
e

(
%
)
Off
On
*
Fig. 4. Means and standard errors of the pedestrian detection rate, as a function of
distance, eccentricity, and whether the car DRLs were on or off (*p<.05).
not-correctly-identied vulnerable road users) by four percentage
points.
In addition to the systematic effect of car DRLs, we also found
that motorcyclists were detected better than pedestrians and
cyclists as a whole. This was probably due to their greater con-
spicuity owing to their slightly bigger size and the presence of their
headlights that were easy to see.
Furthermore, the experiment replicated previous ndings on
the inuence of target distance and eccentricity on detection per-
formance: the detection rate increased by 39 and 32 percentage
points respectively, when the vulnerable road user appeared at
short distance and/or in the center of the picture. Regarding the
effect of distance, this variable determinedthe objects angular size,
which has been shown to be an important conspicuity-inuencing
feature (e.g., Engel, 1971; Jenkins and Cole, 1986). Eccentricity is
also known to affect the detection of objects in a visual scene (e.g.,
Engel, 1971; Rog et al., 2009). In our experiment, the participants
had to initially xate the center of the screen, so no (or only small)
eye movements were required to see and identify the centered tar-
gets when the photograph was displayed. By contrast, detecting
off-centered targets needed larger eye movements and more time.
4.1. Inuences on detection performance
Our discussion will nowlook in greater detail at the effect of car
DRLs on motorcyclist detection. The results demonstrated a sig-
nicant decline in the ability to perceive motorcycles in a car DRL
environment compared to conditions in which the cars headlights
were unlit. The car-DRL effect tended to occur when the motor-
cycle was far away, and when it was located in the center of the
visual scene. The negative effect of car DRLs at greater distances
suggests that the DRLs generated competing light patterns in con-
ditions where the motorcycles were hard to see because of their
small angular size. At shorter distances, where the angular size of
the motorcycle and thus its inherent conspicuity were greater, car
DRLs had little or no impact.
Concerning eccentricity, the adverse effect of car DRLs tended
to show up when the motorcycle was located in the middle of
the picture, which is a favorable condition for target detection.
The ndings suggest that in these conditions, it was especially
motorcycle identication that was hampered by car DRLs: both
motorcycles and cars had their daytime lights on and so motor-
cycles had lost their visual signature as the sole users of DRLs. In
the off-centered position, on the other hand, motorcycle percep-
tion was not affected as much, perhaps due to the very short time
available for exploration. This situationprovedto be highly difcult
(the overall detection rate was only 37%), so the impact of car DRLs
did not turn out to be a decisive factor.
Regarding the cyclists and pedestrians the two types of road
users that did not have lights in our experiment we assumed that
V. Cavallo, M. Pinto / Accident Analysis and Prevention 49 (2012) 7885 83
the masking effect of car DRLs would be even more detrimental for
these users, but our results did not support this hypothesis. To the
same extent as observed for motorcycles, cyclists and pedestrians
were more difcult to perceive when surrounded by cars with their
DRLs on. As a whole, the adverse effect of DRLs was of comparable
magnitude for the three types of vulnerable users.
Unlike motorcyclists, cyclists were less well detected in the
presence of car DRLs when they were closer and in off-centered
locations. For pedestrians, the masking effect of car DRLs was
independent of distance, but was more pronounced when the
pedestrian was located in the center of the scene. These ndings
must be taken with caution, for several reasons. First of all, all types
of vulnerable road users were perceived less well in the presence
of car DRLs, in every distance and eccentricity condition, although
the difference was not always statistically signicant. Secondly, it is
likely that other factors not controlled in our experimental design
may have entered into the picture. While the motorcycles all had a
common visual element, i.e., their headlights were turned on, the
visual appearance of cyclists and pedestrians involved more com-
plex and more varied patterns of colors and contours. In addition,
the visual characteristics of the road environments were diverse.
It is therefore conceivable that for the cyclists and pedestrians, the
impact of car DRLs in the various distance and eccentricity condi-
tions might have beendue to gure/ground effects, whichwe could
not control in this experiment. The importance of these character-
istics has been pointed out in a number of studies (Comelli et al.,
2008; Gershon et al., in press; Olson et al., 1981; Rog et al., 2010),
but our experiment was not designed to examine these factors.
4.2. Methodological considerations
As a whole, our results suggest that car DRLs substantially hin-
der the perception of motorcyclists, and of cyclists and pedestrians
as well. They are not in line with the ndings of most earlier exper-
imental studies. Several reasons can be brought to bear to explain
these discrepancies. Our use of photographs of real trafc scenes
in an urban context allowed us to create realistic, complex, and
unpredictable visual conditions in which a road users conspicuity
was likely to be critical. Furthermore, the brevity of scene exposure
in our study enabled us to simulate situations of rapid information
intake where automobilists had only enough time to glance before
making a decision. Although this strategy is certainly not the pre-
vailing one, it is probably usedinsome cases (whenthe driver is ina
hurry, for example) and is bound to generate the kind of perceptual
errors that cause accidents.
Regarding the validity of the experimental situation studied
here, several aspects of our methodological choices deserve dis-
cussion. Three of these choices the vehicle congurations we
presented, our use of photographs, and the lack of movement in
the visual scenes could be regarded as limitations that reduce the
generalizability of our results.
For the vehicle congurations in the scenes, our experiment
explored only those situations where a vulnerable user was
betweentwocars or next toacar. We didnot examinecertainhighly
critical situations wheretheobserver might confusethemotorcycle
lights with those of an automobile (e.g., the motorcycle is in front
of a car and its headlight coincides with one of the cars headlights)
(Knight et al., 2006). However, we can assume that the car DRL
effect observed in our situations, where it was easier to differenti-
ate between the respective headlights of the motorcycles and the
automobiles, would be even more likely to occur in a more visually
demanding situation.
The advantage of using photographs was that we could study
complex trafc scenes requiring selective attention; the disadvan-
tage was that photographs displayed on a screen cannot render the
trueluminosityof thescenes, nor theactual contrast levels between
the vehicle lights and their environment, as automobilists would
see themin real driving situations. Although we maximized these
contrasts by choosing days withovercast skies, the levels were nev-
ertheless belowthose found in real-world situations. Some studies
have suggestedthat the maskingeffect of car DRLs occurs especially
when the cars headlights are very bright (e.g., Cobb, 1999) and
thus afford a strong contrast with the environment. The fact that
we found a car DRL effect for contrast levels belowthose existing in
realityallows us toexpect asimilar if not greater effect inareal driv-
ing situation. It wouldbe worthwhile to conduct some experiments
in a real situation (on a track, for example) in order to gain further
insight into the roles played by luminosity and contrast for motor-
cycles and automobiles with DRLs. It would be extremely difcult,
however, tocreate complexvisual conditions while simultaneously
controlling vehicle conguration and exposure time.
The lack of movement in our visual scenes is another aspect
that needs to be addressed. Clearly, a moving object in a station-
ary visual scene can be seen as an element of conspicuity that we
did not take into account in our study. This type of situation might
exist, for example, when an automobilist stops at an intersection
and a motorcycle is approaching. Note that the angular movement
of a motorcycle approaching for a short period of time would prob-
ably fall below the threshold of perception, especially when the
motorcycle is coming head on towards the car, which is the most
frequent type of accident situation (head-on collisions) (Brooks
et al., 2005). Moreover, it has been shown that in gap acceptance
situations, motion cues are not always processed by road users as
they make decisions. This seems to be particularly true of seniors,
whose motion perception thresholds are higher (as compared to
young road users), especially for slow movements (Snowden and
Kavanagh, 2006), and whose processing speed is lower (Salthouse,
1996). In a street-crossing situation, for example, we found that
elderly pedestrians failed to make use of motion cues and relied
instead on heuristics based on the approaching vehicles distance
(Lobjois and Cavallo, 2007). This strategy has also been observed
among young pedestrians whowere givena limitedamount of time
to decide whether to cross the street (Lobjois and Cavallo, 2007).
In cases where there is just enough time to glance at the scene,
as in our experiment where display time was short, it thus seems
unlikely that motion information could be taken into account.
If we agree that our experiment was effective at recreating a cer-
tain number of real driving situations that were critical in terms of
the sensorial conspicuity of motorcycles, then howcan we account
for the fact that the adverse car DRL effect we observed here on the
perception of motorcycles and other vulnerable users, has hardly
ever been found in accident studies?
One reason might lie in howthe assessments were made in the
various studies. In certain analyses on car DRL effects, accidents
involving unprotected users are not taken properly into account.
Another reason could be that when automobilists use DRLs, other
road users such as motorcyclists, cyclists, and pedestrians can see
the cars better and avoid collisions with them. In line with this,
some studies suggest that car DRLs are also benecial to vulner-
able users (Bergkvist, 2001; NHTSA, 2000; Rumar, 2003; Tessmer,
2004; Thompson, 2003). It is assumed, for example, that car DRLs
promote the detection of vehicles in the peripheral visual eld
(Andersson and Nilsson, 1981) and are particularly benecial for
older users whose useful visual eld and contrast sensitivity are
reduced (Rumar, 2003).
For all accidents taken together, then, the increased visibility
of automobiles for vulnerable users could compensate for the
decreased visibility of vulnerable users for automobilists, and
could even improve the overall safety of these users. The ndings
of our study suggest, however, that vulnerable road-user safety can
be further improved by helping automobile drivers better perceive
these users in a car DRL environment. Increasing vulnerable-user
84 V. Cavallo, M. Pinto / Accident Analysis and Prevention 49 (2012) 7885
conspicuity could indeed help automobilists detect these users
better and thereby avoid accidents caused by failure to yield the
right of way. As we have seen, this type of accident accounts for
two thirds of the collisions between cars and motorcycles.
4.3. Further research
Although the globally positive safety effect of car DRLs is
acknowledged, our study suggests that more attention should be
paid to motorcyclists and other vulnerable road users when intro-
ducing car DRLs.
An important issue for future research concerns the ergonomics
of motorcycle headlights. The goal of such research could be to nd
a new, unique visual signature for motorcycles that would make
them easier to detect and identify. A number of studies in this
area are already in progress in Europe (Rssger et al., in press),
Israel (Gershon et al., in press), Japan (Maruyama et al., 2009;
Tsutsumi and Maruyama, 2008) and the United States (Binder et al.,
2005). Light congurations like the ones proposed by Tsutsumi and
Maruyama (2008), and Maruyama et al. (2009), which increase
the visual width and height of motorcycles, could not only pro-
motemotorcycledetectabilitybut alsoimproveestimations of their
speed, distance, and time-to-arrival. Another way of enhancing
conspicuity is to change the visual features of motorcycling gear
(helmets, jackets, vests) in view of making motorcycles more vis-
ible to automobilists (Comelli et al., 2008; Olson et al., 1981). This
approach could also prove useful for enhancing the visibility of
bicycle riders. In France, for example, a number of city bikers have
started wearing uorescent yellowvests, and it would be interest-
ing to assess their impact in terms of safety and behavior. To our
knowledge, no experimental or accident data is available on this
subject.
Another researchtrendthat has emergedlooks intoways of rais-
ing the cognitive conspicuity of motorcycles, not only by improving
the driving skills of automobilists but also by changing their atti-
tudes. A number of authors have stressed that the low sensorial
conspicuity of motorcycles cannot fully account for the large num-
ber of automobile-motorcycle collisions. They suggest that the
non-perception of motorcycles might also be linked to unrealistic
expectations on the part of car drivers (Brooks and Guppy, 1990;
Hancock et al., 1990). In this vein, it has been shown that automo-
bilists licensed to drive a motorcycle (Magazz et al., 2006) as well
as persons with a motorcycle rider among the family or who them-
selves have beena motorcycle passenger (Brooks and Guppy, 1990)
have fewer accidents with motorcycles than other automobilists
do. Implementing road-safety measures based on driver training
andeducationcouldthus be anadditional way of improving motor-
cycle perception and motorcyclists safety.
Acknowledgements
The authors would like to thank Jennifer Oxley and Michael
Sivak for their helpful comments on earlier versions of this paper.
We also express our gratitude to Vivian Waltz for the transla-
tion, Jean-Philippe Cavallo for the computer graphics, and Karoline
Rbisch for her help in preparing the experimental material and
running the experiment. This project was nanced by the European
research project entitled 2-Wheelers Behavior and Safety (2-be-
safe) (7th FP proposal).
References
ACEM, 2009. In-depth investigations of accidents involving powered two
wheelers (MAIDS). Retrieved August 28, 2010, from: http://www.maids-
study.eu/pdf/MAIDS2.pdf.
Andersson, K., Nilsson, G., 1981. The Effects on Accidents of Compulsory Use of
Running Lights During Daylight in Sweden. VTI, Linkping, Sweden.
Ball, K., Owsley, C., Sloane, M.E., Roenker, D.L., Bruni, J.R., 1993. Visual attention
problems as a predictor of vehicle crashes in older drivers. Investigative Oph-
thalmology and Visual Science 34, 31103123.
Bergkvist, P., 2001. Daytime running lamps (DRLs) a North American success story.
In: Proceedings of the 17th International Technical Conference on the Enhanced
Safety of Vehicles, Amsterdam, The Netherlands, Paper No. 395.
Binder, S., Perel, M., Pierowicz, J., Gawron, V., Wilson, G., 2005. Motorcycle con-
spicuity and the effects of motor vehicle eet Daytime Running Lights (DRLs).
In: Proceedings of the 6th International Symposium on Automotive Lighting
(ISAL), Darmstadt, Germany.
Brenac, T., Clabaux, N., Perrin, C., Van Elslande, P., 2006. Motorcyclist conspicuity-
related accidents in urban areas: a speed problem? Advances in Transportation
Studies 8, 2329.
Brendicke, R., Forke, E., Schfer, D., 1994. Auswirkungen einer allgemeinen Tages-
lichtpicht auf die Sicherheit motorisierter Zweirder. In: 6. Fachtagung
Motorrad, VDI Berichte 1159, Dsseldorf, Germany, pp. 283318.
Brooks, A.M., Chiang, D.P., Smith, T.A., Zellner, J.W., Peters, J.P., Compagne, J.,
2005. A driving simulator methodology for evaluating enhanced motorcycle
conspicuity. In: Proceedings of the 19th International Technical Confer-
ence on the Enhanced Safety of Vehicles, Washington, DC, Paper No.
05-0041-O.
Brooks, P., Guppy, A., 1990. Driver awareness and motorcycle accidents. In: Proceed-
ings of the International Motorcycle Safety Conference, Orlando, Florida.
Brouwer, R.F.T., Jansen, W.H., Theeuwes, J., Duistermaat, M., Alferdinck, J.W.A.M.,
2004. Do Other Road Users Suffer from the Presence of Cars that have their
Daytime Running Lights On? TNO, Soesterberg, The Netherlands.
Cairney, P., Styles, T., 2003. Review of the Literature on Daytime Running Lights
(DRLs). Australian Transport Safety Bureau, Canberra, Australia.
Cobb, J., 1992. Daytime Conspicuity Lights. Transport Research Laboratory,
Crowthorne, Berkshire, United Kingdom.
Cobb, J., 1999. Vehicle Lighting and Signalling. Transport Research Laboratory,
Crowthorne, Berkshire, United Kingdom.
Comelli, M., Morandi, A., Magazzu, D., Bottazzi, M., Marinoni, A., 2008. Brightly
coloured motorcycles and brightly coloured motorcycle helmets reduce the
odds of a specic category of road accidents: a case-control study. Biomedical
Statistics and Clinical Epidemiology 2 (1), 18.
Connors, M.M., 1975. Conspicuity of Target Lights: The Inuence of Color. NASA,
Washington, DC.
Crundall, D., Crundall, E., Clarke, D., Shahar, A. Why do car drivers fail to give
way to motorcycles at T-junctions? Accident Analysis and Prevention, in press,
doi:10.1016/j.aap.2010.10.017.
Crundall, D., Humphrey, K., Clarke, D., 2008. Perceptionandappraisal of approaching
motorcycles at junctions. TransportationResearchPart F: Trafc Psychology and
Behaviour 11 (3), 159167.
Dahlstedt, S., 1986. A Comparison of Some Daylight Motorcycle Visibility Treat-
ments. VTI, Linkping, Sweden.
EC, 2009. Motorcycles, mopeds and road safety. Retrieved December 14, 2009, from
EC Web site: http://ec.europa.eu/transport/road safety/users/motorcyclists-
mopdeds/index en.htm.
Elvik, R., 1996. A meta-analysis of studies concerning the safety effects of
daytime running lights on cars. Accident Analysis and Prevention 28 (6),
685694.
Engel, F.L., 1971. Visual conspicuity, directed attention and retinal locus. Vision
Research 11 (6), 563575.
Engel, F.L., 1974. Visual conspicuity and selective background interference in eccen-
tric vision. Vision Research 14 (7), 459471.
Engel, F.L., 1977. Visual conspicuity, visual search and xation tendencies of the eye.
Vision Research 17 (1), 95108.
Farmer, C.M., Williams, A.F., 2002. Effects of daytime running lights on multiple-
vehicle daylight crashes in the United States. Accident Analysis and Prevention
34 (2), 197203.
Federal Ofce of Road Safety, 1997. Vehicle type and risk of travelling on the road.
In: Monograph, vol. 17. Canberra, Australia.
FEMA, 2006. Saving [car drivers] lives with daytime running lights: Consultation
paper. FEMAs comments. EC Web site, Brussels, Belgium.
Gershon, P., Ben-Asher, N., Shinar, D. Attention and search conspicuity of motorcy-
cles as a function of their visual context. Accident Analysis and Prevention, in
press, doi:10.1016/j.aap.2010.12.015.
Guyot, R., 2008. Gisements de Scurit Routire: Les Deux-roues Motoriss. La Doc-
umentation Franc aise, Paris, France.
Hancock, P.A., Wulf, G., Thom, D., Fassnacht, P., 1990. Driver workload during differ-
ing driving maneuvers. Accident Analysis and Prevention 22 (3), 281290.
Hendtlass, J., 1992. Appendix C: Literature review. Daytime running lights. In: Social
Development Committee Inquiry into Motorcycle Safety in Victoria. Parliament
of Victoria, Melbourne, Australia.
Hrberg, U., Rumar, K., 1979. The effects of running lights on vehicle conspicuity in
daylight and twilight. Ergonomics 22 (2), 165173.
Horswill, M.S., Helman, S., Ardiles, P., Wann, J.P., 2005. Motorcycle accident risk
could be inated by a time to arrival illusion. Optometry and Vision Science 82
(8), 740746.
Hughes, P.K., Cole, B.L., 1984. Search and attention conspicuity of road trafc control
devices. Australian Road Research Board 14 (1), 19.
Hurt, H.H.J., Ouellet, J.V., Thom, D.R., 1981. Motorcycle accident cause factors and
identication of countermeasures, vol. 1. NHTSA, Washington, DC.
IRTAD, 2009. Road safety. Annual report. Retrieved August 28, 2010. from:
http://www.internationaltransportforum.org/irtad/pdf/09IrtadReportFin.pdf.
V. Cavallo, M. Pinto / Accident Analysis and Prevention 49 (2012) 7885 85
Itti, L., Dhavale, N., Pighin, F., 2003. Realistic avatar eye and head animation using
a neurobiological model of visual attention. In: Proceedings of the Applications
and Science of Neural Networks, Fuzzy Systems, and Evolutionary Computation
VI, San Diego, CA, pp. 6478.
Itti, L., Koch, C., 2000. A saliency-based search mechanismfor overt and covert shifts
of visual attention. Vision Research 40, 14891506.
Janoff, M.S., Cassel, A., 1971. Effect of daytime motorcycle headlight laws. On motor-
cycle accidents. Highway Research Record 377, 5363.
Jenkins, S.E., Cole, B.L., 1986. Daytime conspicuity of road trafc control devices.
Transport Research Record 1093, 7480.
Knight, I., Sexton, B., Bartlett, R., Barlow, T., Latham, S., Mccrae, I., 2006. Daytime
running lights (DRLs): a review of the reports fromthe European Commission.
TRL, Crowthorne, Berkshire, United Kingdom.
Koornstra, M., Bijleveld, F., Hagenzieker, M.P., 1997. The Safety Effects of Daytime
Running Lights. SWOW, Leidschendam, The Netherlands.
Lobjois, R., Cavallo, V., 2007. Age-relateddifferences instreet-crossing decisions: the
effects of vehicle speed and time constraints on gap selection in an estimation
task. Accident Analysis and Prevention 39 (5), 934943.
Magazz, D., Comelli, M., Marinoni, A., 2006. Are car drivers holding a motorcycle
licence less responsible for motorcycle-car crash occurrence? A non-parametric
approach. Accident Analysis and Prevention 38 (2), 365370.
Maruyama, K., Tsutsumi, Y., Murata, Y., 2009. Face design, lighting systemdesign for
enhanced detection rate of motorcycles. Journal of the Society of Automotive
Engineers of Japan October.
Muller, A., 1984. Daytime headlight operation and motorcyclist fatalities. Accident
Analysis and Prevention 16 (1), 118.
NHTSA, 2000. A preliminary assessment of the crash-reducing effectiveness of pas-
senger car daytime running lamps (DRLs). Washington, DC.
NHTSA, 2008. Trafc safety facts. Washington, DC.
Olson, P.L., Halstead-Nussloch, R., Sivak, M., 1981. The effect of improvements in
motorcycle/motorcyclist conspicuity on driver behavior. Human Factors 23 (2),
237248.
Pai, C.-W., 2011. Motorcycle right-of-way accidents a literature review. Accident
Analysis and Prevention 43, 971982.
Paine, 2003. A Review of Daytime Running Lights. NRMA Motoring & Services and
RACV, Sydney, Australia.
Ramsey, J.D., Brinkley, A., 1977. Enhanced motorcycle noticeability through daytime
use of visual signal warning devices. Journal Safety Research 9 (2), 7784.
Riemersma, J.B.J., Wertheim, A.H., Welsh, M., Bakker, P.J., 1987. Deopvallendheidvan
het attentielicht. Institute for Perception, TNO, Soesterberg, The Netherlands.
Rssger, L., Hagen, K., Krzywinski, J., Schlag, B. Recognisability of different congura-
tions of front lights on motorcycles. Accident Analysis and Prevention, in press,
doi:10.1016/j.aap.2010.12.004.
Rog, J., Ferretti, J., Devreux, G., 2010. Sensory conspicuity of powered two-wheelers
during ltering manuvres, according to the age of the car driver. Le Travail
Humain 73 (1), 730.
Rog, J., Otmani, S., Bonnefond, A., Pbayle, T., Muzet, A., 2009. Effect of a short nap
on the alertness of young drivers: repercussion on the perception of motorcy-
cles according to extent of the useful visual eld of the driver. Transportation
Research Part F: Trafc Psychology and Behaviour 12 (2), 143154.
Rumar, K., 2003. Functional Requirements for Daytime Running Lights. University
of Michigan Transportation Research Institute, Ann Arbor, MI.
Salthouse, T.A., 1996. The processing-speed theory of adult age differences in cog-
nition. Psychological Review103 (3), 403428.
Schnebeck, S., Ellmers, U., Gail, J., Krautscheid, R., Tews, R., 2005.
Abschtzung mglicher Auswirkungen von Fahren mit Licht am Tag (Tag-
fahrleuchten/Abblendlicht) in Deutschland. Bundesanstalt fr Strassenwesen,
Bergisch Gladbach, Germany.
Shahar, A., van Loon, E., Clarke, D., Crundall, D. Attending overtaking cars and
motorcycles through the mirrors before changing lanes. Accident Analysis and
Prevention, in press, doi:10.1016/j.aap.2011.01.001.
Snowden, R.J., Kavanagh, E., 2006. Motion perception in the ageing visual system:
minimummotion, motion coherence, and speed discrimination thresholds. Per-
ception 35 (1), 924.
Tessmer, 2000. A preliminary assessment of the crash-reducing effectiveness of
passenger car daytime running lamps (DRLs). NHTSA, Washington, DC.
Tessmer, J.M., 2004. An assessment of the crash-reducing effectiveness of passenger
vehicle daytime running lamps (DRLs). NHTSA, Washington, DC.
Theeuwes, J., Riemersma, J., 1995. Daytime running lights as a vehicle collision
countermeasure: the Swedish evidence reconsidered. Accident Analysis and
Prevention 27 (5), 633642.
Thompson, P., 2003. Daytime Running Lamps for Pedestrian Protection. Society of
Automotive Engineers, Warrendale, PA.
Thomson, G.A., 1980. The role frontal motorcycle conspicuity has in road accidents.
Accident Analysis and Prevention 12 (3), 165178.
Tofemire, T.C., Whitehead, P.C., 1997. An evaluation of the impact of daytime run-
ning lights ontrafc safety inCanada. Journal of Safety Research28(4), 257272.
Tsutsumi, Y., Maruyama, K., 2008. Long lighting systemfor enhanced conspicuity of
motorcycles. JSAE Transaction 39 (5), 119124.
Van Elslande, P., Jaffard, M., 2010. Typical human errors in trafc accidents involving
powered two-wheelers. In: Proceedings of the 27th International Congress of
Applied Psychology, Melbourne, Australia, pp. 13631364.
Wang, J.-S., 2008. The Effectiveness of Daytime Running Lights for Passenger Vehi-
cles. NHTSA, Washington, DC.
Williams, M.J., Hoffmann, E.R., 1979. Motorcycle conspicuity and trafc accidents.
Accident Analysis and Prevention 11 (3), 209224.
Woltman, H.L., Austin, R.L., 1974. Some day and night visual aspects of motorcycle
safety. Transportation Research Record 502, 18.
Wulf, G., Hancock, P.A., Rahimi, M., 1989. Motorcycle conspicuity: an evaluation and
synthesis of inuential factors. Journal of Safety Research 20 (4), 153176.
Yee, H., Pattanaik, S.N., Greenberg, D.P., 2001. Spatio-temporal sensitivity and visual
attention for efcient rendering of dynamic environments. ACM Transactions
on Graphics 20 (1), 3965.
Zador, P.L., 1985. Motorcycle headlight-use laws and fatal motorcycle crashes in the
US, 197583. American Journal of Public Health 75 (5), 543546.
Zimmermann, P., Fimm, B., 2002. A test battery for attentional performance. In:
Leclercq, M., Zimmermann, P. (Eds.), Applied Neuropsychology of Attention.
Theory, Diagnosis and Rehabilitation. Psychology Press, London, pp. 110151.