Nighttime Legibility of Highway Signs Alicia Bower, Patrick Hamiel, Elizabeth Overmoe, Beth Peters,

1
Nighttime Legibility of Highway SignsAlicia
Bower, Patrick Hamiel, Elizabeth Overmoe, Beth
Peters, Elijah Stines

• ABSTRACT
• The nighttime legibility of highway signs is a
  very important aspect of driver safety. Yet,
  very little research has been done in this area.
  In fact, there are no official laws or consensus
  guidelines regarding the luminance levels of road
  signs. This study attempts to find the level of
  luminance needed for optimal legibility during
  nighttime driving. Two types of signs were
  tested. All were 3M brand retroreflective
  material, one type at 15 reflectivity, another
  at 100 reflectivity (which changed the light
  distribution by moving the peak luminance closer
  to the sign). Our study set out to find the
  difference in legibility levels between each type
  of sign, the difference in legibility of the
  signs at a static and a dynamic viewing
  condition, and the interaction of sign type and
  viewing condition. Results of this study will
  provide important information for future planning
  of highway sign production. More specifically,
  our results will be used to establish minimum
  required luminance values as well as determine if
  drivers can benefit from changes in the spatial
  luminance distribution of light reflected from
  highway signs at night.

RESULTS Our analysis of the data looked that the
difference in legibility distances for each type
of sign and the difference between the two
viewing conditions. We used a 2x2 factor ANOVA
test to analyze the data. The results found that
there was a significant difference in legibility
distances between the 15 and 100
retroreflective signs, F(1,9)20.6, plt 0.001.
The mean legibility distance for the 15
reflectance signs was 91.9 m with a standard
error of 9.1 m. The mean distance for the 100
reflectance signs was 104.2 m with a standard
error of 9.1 m. There was also a significant
difference between the static and dynamic driving
condition, F(1,9)48.6, plt0.000. The mean
legibility distance for the static viewing
condition was 114.9 m with a standard error of
8.0 m. The mean distance for the dynamic
condition was 81.2 m with a standard error of
10.7 m. These results are illustrated in
Figures 2 and 3, respectively. With an average
legibility difference of 33.7 m and an average
speed of 26.5 m/sec at the critical response
point, the inferred perception response time
(PRT) was 1.28 seconds.
INTRODUCTION How well people see objects during
the day is determined by an objects color and
the amount of light that it gives off compared to
the amount of light given off by its
surroundings. At night, people use the light
emitted from a cars headlights to view highway
signs. The more reflective a road sign is, the
more light that will be emitted back towards the
driver. The Manual on Uniform Traffic Control
Devices (MUTCD) requires that highway signs be
made of retroreflective material. However, the
manual does not give minimum values for the level
of retroreflectivity required. Current research
is focusing on designated minimum values of
reflectivity for road signs. The results will be
used to decide when and what signs need to be
replaced (FHWA RD-97-074). The luminance of a
sign depends on how the material is constructed
and on the amount of light being shed on the
sign. Retroreflective materials use glass beads
or cube corner elements that reflect light
directly back to where it came from. Whereas,
mirror reflection reflects light at a 90 degree
angle to where it came from. Highway statistics
reveal that accidents are three times more likely
to happen at night than during the day. Reduced
visibility is believed to play a major role in
this statistic. To minimize reduced visibility,
it is important to know how different levels of
reflection affect a persons ability to read
highway signs. To define minimum levels of
retroreflectivity, we must know how signs are
viewed at the static driving condition (0-6 mph)
and at the dynamic condition (roadway speed).
This difference, as well as the type of material,
can have an impact on the viewing distances.
This study was conducted at a test track on the
3M plant in Cottage Grove, MN. A picture of this
track is shown in Figure 1. 3M is the leading
developer of state-of-the-art retroreflective
material.
Figure 2. Legibility Distance as a function of
Sign Reflectance
DISCUSSION Our results show a sign with 100
reflectance is legible from a greater distance
than a sign with 15 reflectance. The 100
reflectance signs were read an average of 12.3 m
earlier than the 15 reflectance signs. A
difference of 12.3 m can make a big difference in
the time a driver has to make a decision. The
use of 100 reflectance signs could certainly
increase highway safety. Since there is a
significant decrease 33.7 m in the legibility
distance of signs from the static to the dynamic
viewing condition, it seems that it would greatly
benefit highway drivers if 100 reflectance signs
were used. The PRT lets us know how long it
takes the driver to read the sign from the point
that they have enough visual information to read
the sign at the static viewing condition to where
the driver actually reads the sign at the dynamic
condition. This can help researchers decide
where to place signs if they know by what point
on the road the driver must have the information
given on the sign.
METHOD The study required the use of ten
subjects. Each were utilized as both
experimenters and participants. They were given
the task of completing ten driving laps around a
track at the 3M Company, located in southern
Minneapolis, after dark. The first round
consisted of a practice run allowing the
participants to become comfortable with their
driving environment. The following nine runs
around the track were a sequence of either three
static and six dynamic or three dynamic and three
static. The sequence was randomly chosen for
each subject, yet the study was conducted so that
half of the participants received one way, while
the remaining half received the latter. The signs
viewed by each subjects were also randomly
chosen. Half of the signs were 15 reflectivity
(low luminance) and the other half at 100
reflectivity (high luminance). The static run
was conducted at a mere five mile an hour speed,
and it consisted of viewing two signs along the
course of the track. The subject was instructed
to loudly and quickly read the sign aloud
immediately after recognition. The dynamic run
included viewing one sign on the track at a speed
of approximately 55 miles per hour. The subject
was, once again, instructed to read the sign
aloud immediately upon recognition. The test car
is equipped with a video system that records both
the response of the participant and the distance
at which the sign was recognized. The videos
were then scored by the students upon returning
to the University of South Dakota. This data was
scored on, first, what distance the signs were
recognized, and secondly, the correlation those
numbers had with the type of material with which
the sign was made. Very few materials were needed
for this research. The test car was a Toyota
Avalon, and inside of it were both the video
camera that was recording all of the information,
along with a lap top that contained the computer
programs necessary to compile and make sense of
all the data. Along the track there formed two
stations. At each of these locations were a
variety of signs that were continuously being
placed on holders, which were sustained by
sandbags to keep them from falling. These signs
were the same height as domestic highway signs,
in order to eliminate variance between the study
and FHWA regulations. Walkie talkies were
necessary at each station to ensure communication
was upheld between those seated in the car and
those placing signs. Once all information was
compiled and brought back to the University of
South Dakota, it was scored through a computer
system in the visual lab of the psychology
department.
Figure 3. Legibility Distance as a function of
Viewing Condition
CONCLUSION This study shows that retroreflective
light patterns that are returned from highway
signs must be reevaluated to factor in the
drivers need for sign legibility at nearer
distances. Past planning of production and
placement of highway signs been based on
legibility distances determined from the static
viewing condition (100 m). Our findings show
that, when at high speeds, a driver needs more
time to process the information from a sign.
Therefore, when determining the placement of
highway signs, one must take into account the
reduction of legibility at a dynamic driving
condition. Our results also show that high
levels of retroreflective sheeting could be used
to increase driver safety. 100 retroreflective
signs can be read from further distances than the
15 retroreflective signs, thereby giving the
driver more time to process the information and
take the appropriate actions.
Figure 3. Birdseye view of the 3M Test Track