Conference PaperPDF Available

Abstract and Figures

Rain drastically modifies the visual environment of road users, particularly at night. It changes visibility through its effects on headlights, windshield, pavement and markings. Rain lessens the performance of headlamps and other light sources by filtering part of their luminous power, thus reducing the illuminance on the roadway ahead of the vehicle. Rain affects the capacity of the driver to see through the windshield. Rain also affects visibility by changing the amount of headlight retro-reflected by the road surface toward the driver. The film of water on the pavement makes delineation and pedestrian crossing markings almost invisible by cancelling the retroreflective properties of the beads in the painting materials. The same physical phenomenon makes the pavement appear darker than in dry conditions. This is a brief list of commonplace facts on the visual effects of rain (10). In the following, we seek to explicit the physical and psychophysical ground of these facts, based on scientific references when available. The outline of the paper is as follows: in the first section, we study the visual effects of the falling rain; in the second section, we tackle the visual effects of sprayed water; in the third section, we list the properties of wet materials, specially pavement markings; finally, we propose synthetic diagrams describing the effects of rain on roadway visibility.
Content may be subject to copyright.
Review of the Mechanisms of Visibility Reduction by
Rain and Wet Road
Nicolas Hautière, Eric Dumont, Roland Brémond, Vincent Ledoux
Keywords: rain, wet road, visibility, windshield
1 Introduction
Rain drastically modifies the visual environment of road users, particularly at
night. It changes visibility through its effects on headlights, windshield, pavement
and markings. Rain lessens the performance of headlamps and other light
sources by filtering part of their luminous power, thus reducing the illuminance on
the roadway ahead of the vehicle. Rain affects the capacity of the driver to see
through the windshield. Rain also affects visibility by changing the amount of
headlight retro-reflected by the road surface toward the driver. The film of water
on the pavement makes delineation and pedestrian crossing markings almost
invisible by cancelling the retroreflective properties of the beads in the painting
materials. The same physical phenomenon makes the pavement appear darker
than in dry conditions.
This is a brief list of commonplace facts on the visual effects of rain [10]. In the
following, we seek to explicit the physical and psychophysical ground of these
facts, based on scientific references when available.
The outline of the paper is as follows: in the first section, we study the visual
effects of the falling rain; in the second section, we tackle the visual effects of
sprayed water; in the third section, we list the properties of wet materials,
specially pavement markings; finally, we propose synthetic diagrams describing
the effects of rain on roadway visibility.
2 Visual effects of the falling rain
2.1 The nature and microstructure of rain
Rain is a population of water droplets falling, interacting with each other and with
the environment. While falling, a rain drop undergoes rapid shape distortions.
This shape is size-dependent. Small drops are usually spherical, but as their size
grows, they tend to a spherical oblate shape. The shape of a rain drop is
described in [2] as the cosine distortion of a sphere at the tenth order:
( )
θ θ
 
= +
 
 
( ) 1 cos
r a c n (1)
where a is the radius of the undistorted sphere, c
are coefficients which
depend on the radius of the drop, and θ is the elevation polar angle. θ=0
corresponds to the direction of the rain. Shapes of various sized drops are
presented in Figure 1a. Rain drops come in a wide range of sizes. Their size
distribution is often modeled using Marshall-Palmer distribution [16]:
( )
N a N e
where a is the radius of a drop, N(a) the number of drops per volume unit with
sizes between a and a+da, N
=0.08 cm
, Λ=82R
and R is the rain density in
. This distribution is plotted in Figure 1b.
(a) (b)
Figure 1: (a) shapes of rain drops. (b) Marshall-Palmer rain drop size distribution.
Rain drops fall at a constant speed called the terminal velocity. An empirical
study is presented in [11] on the terminal velocity of rain drops for different drop
sizes. This data are approximated in [27] with the following function:
− ×
= − 3 1,15
1,57 10
9,4 1
V e (3)
2.2 Light scattering in rain
Some experiments were conducted in an attempt to relate optical extinction on a
long distance to rain density. [26] reports the results from five other
investigations, and concludes that the optical depth τ obtained from the
measurements corresponds within 25% to the value computed on the basis of
different rain drop distributions. The optical depth is computed by integrating the
extinction coefficient k
along the optical path L as follows:
k dz
The general relation found between the extinction coefficient k
) and rain
intensity R (mm.h
) is the following:
k aR
differ with respect to the location and the optical devices used in
the experiments. Experimental curves are plotted in Figure 2.
Finally, [21] measured the back-scattering of a light source in rain. Based on
these measurements, they proposed an empirical model. However, the relevance
of this model was not tested since.
Figure 2: Different experimental curves relating the atmospheric extinction coefficient and the
intensity of the rain.
2.3 Consequences on roadway visibility
Light scattering in rain is rather limited. Using an analogy with the visual effects
of fog, the effects of scattering in rain can be a problem for driving when the
meteorological visibility (V
) falls below 400 m, which is equivalent to a
300 mm.h
rain according to Eq. (5). Such levels of precipitation are seldom
3 Visual effects of sprayed water
3.1 Visual effects of rain on the windshield
To the best of our knowledge, there is no analytic model for the overall reduction
of visibility induced by rain on the windshield. [9] focus on the appearance of rain
drops. They show that the field of view refracted by a spherical drop is about
165°, and assimilate the drop with a fish-eye lens. The corresponding optical
diagram is presented in Figure 3a.
(a) (b)
Figure 3: Optical diagram illustrating the fish-eye lens effect created by a rain drop.
One can assume rain drops on the windshield to have a similar effect, save for
the deformation of the spherical drops on the windshield. The drops on the
windshield roughly reflect the road environment, which is illustrated in Figure 3b.
On the other hand, several experimental studies on wiper usage focused on
object visibility and seeing distance. Some driver visibility studies were restricted
to stationary vehicles in artificial rain [13][18]. Other studies were conducted in
actual rain. They showed that wipers do not interfere with the perception of the
road scene with respect to saccadic eye movements [6]. Moreover, seeing
distances are significantly reduced when rain intensity increases [3][12][17]. In
particular, [3] investigated the visibility distance of target vehicles under natural
downpours. The observers were onboard a vehicle, and notified when they
detected a target car while their wipers were engaged or recently stopped. These
experiments showed that detection distances decrease significantly with ambient
lighting and that visibility distance decreases as rain intensity increases. Visibility
distance was found to be lower for observers onboard a moving vehicle (vs.
stationary) because of the higher concentration of water on the windshield.
Based on these experiments, a model was proposed for the visibility distance of
cars through the windshield in rain condition in daytime condition. This model can
be simplified by:
c L
D c rt e
= (6)
where c
, c
, c
are strictly positive constant values, rt characterizes the
accumulation of rain water on the windshield, r being the intensity of the rain and
t the time between wiper movements, and L
the background luminance.
3.2 Visual effects of water sprayed by other vehicles
The water sprayed by vehicles has undeniable effects on visibility. [8] studied
these effects. Unfortunately, no model came out of it, because of several
experimental difficulties which impeded the identification of prevailing
parameters. Other studies showed that splash and spray is reduced by 95% on
porous asphalt compared to other ordinary pavement surfaces. Figure 4, taken
from [20], shows a heavy vehicle on a road section with and without porous
asphalt. However, such a figure is dubious since it is not backed up with a
measurement protocol. The most rigorous works have been conducted for the
development of heavy vehicle spray reduction devices. Even though these
researches do not directly address driver visibility, the metering systems which
were used to study such devices might be used to investigate this particular
problem. A synthesis of the works conducted before 2000 is proposed in [15].
Figure 4: Water sprayed by a heavy vehicle on ordinary and porous asphalt [20].
4 Light reflections on wet materials
4.1 Water at the surface
The water on a surface (e.g. a puddle on the pavement) makes it specular
because of the smooth air-water interface. Optical interactions on such a surface
are governed by Fresnel equation for dielectric materials:
θ θ
1 1 2 2
sin sin
n n (5)
A film of water on a Lambertian surface can also make the surface appear
darker [14]. This is mainly caused by internal reflections at the water-air interface.
Part of the light reflected by the Lambertian surface is reflected back when it hits
the water-air interface. This light is again subject to absorption by the material of
the surface before being reflected again. This leads to a sequence of absorptions
which darkens the surface.
(a) (b)
Figure 5: Illustration of the Fresnel equation: a material with a layer of water on its surface reflects
less light because of internal reflections at the water-air interface.
4.2 Water underneath the surface
The presence of water underneath a surface is another important factor
influencing the appearance of the material. In the case of pulverulent materials,
(sand or limestone), water can penetrate inside holes formerly filled with air. This
modifies the reflection properties of the material, favoring forward scattering [25].
The main reason is that the refraction index of water is higher than the index of
the air, and usually closer to the index of the material. This means that a ray of
light entering the material is less refracted because the refraction index is more
homogeneous when the material is wet. As illustrated in Figure 6, the
consequence is that the ray undergoes more scattering before leaving the
surface. This increases the total amount of absorbed light, and the overall effect
is a material with reduced reflectivity.
Figure 6: Shortest path for a ray of light to enter and exit the material with (left) a 90° mean
scattering angle and (right) a 30° mean scattering angle.
4.3 Consequences on roadway visibility
Rain changes the visual aspect of the road. The road surface appears more
specular or darker, depending on the observation angle. This can be dazzling for
the driver, especially in daytime with the sun at grazing angles, or at night with
opposing headlights. With visual performance impaired by glare, it is more
difficult for the driver to detect hazards. The visibility of retro-reflective road
markings is also particularly impaired. These markings are designed to send
headlight back toward the vehicle. They are usually made of a painting onto
which glass beads with a high refraction index (between 1.5 and 2.5) are
encrusted (Figure 7a). The optical properties of the beads are described in more
details in [29] and [23]. In daytime, on wet roads, retroreflective materials reflect
sunlight, and sometimes appear darker than the pavement. At night, when the
road is slightly wet, the retroreflective efficiency of the beads is reduced, as
illustrated in Figure 7b. When the road is wet and the water layer is higher than
the size of the beads, headlight is mostly reflected at the air-water interface
(Figure 7c), so markings may disappear. This is why all weather markings were
developed. A nice introduction to this particular issue is proposed in [5].
(a) (b) (c)
Figure 7: Optical mechanisms governing retroreflection on pavement markings with encrusted
beads in (a) dry, (b) humid and (c) wet conditions.
5 Conclusion
In this paper, we have provided physical explanations for the effects of rain on
roadway visibility. These explanations are based on either optical or
psychophysical models. We have classified the visual effects of rain into three
main categories. The first category concerns light scattering by rain drops. The
second category concerns splash and spray. It encompasses both rain falling on
the windshield and water splashed by other vehicles. The third category
concerns wet road surfaces, especially road markings, whose appearance is
modified by the water layer.
Figure 8: Visual effects of rain in daytime
In the end, the reduction of visibility caused by rain and sprayed water results
from a combination of these three categories of effects. We can estimate a priori
that the effects of the second and third categories have the highest impact on
visibility, scattering effects being negligible for common intensity downpours. To
give a schematic view of these effects, we propose two diagrams. Figure 8
shows the various effects of rain on daytime visibility: reduced transmission,
atmospheric veil, wet windshield, spray and specular reflections. Figure 9 shows
the nighttime rain situation, with the same effects as in daytime plus specular
reflection overcoming retroreflection on the pavement.
From this review of the literature, we have seen that the mechanisms of visibility
reduction by rain and wet road are numerous. The technical solutions to enhance
the perception of the driver in rainy weather are thus also numerous: adaptive
wipers, adaptive headlights, anti-splash and spray devices, porous pavement,
retroreflective markings… A next step should be to focus on headlights. The
quantitative visibility models should enable to define scenarios and to compute
the necessary power to compensate for the visibility loss, or to find alternative
strategies to compensate for the backscattering of light.
Figure 9: Nighttime visibility (a) in clear weather (b) in rainy weather.
6 References
[1] W. Bachman, T. Wingert, and C. Bassi. Driver contrast sensitivity and
reaction times measured through a salt-covered windshield. Optometry - Journal
of the American Optometric Association, 77(2):67–70, 2006.
[2] K. V. Beard and C. Chuang. A new model for the equilibrium shape of
raindrops. Journal of the Atmospheric Sciences, 44(11):1509–1524, June 1987.
[3] V. Bhise, J. Meldrum, L. Forbes, T. Rockwell, and E. McDowell. Predicting
driver seeing distance in natural rainfall. Human Factors, 23(6):667–682, 1981.
[4] S. G. Bradley, C. D. Stow, and C. A. Lyuch-Blosse. Measurements of rainfall
properties using long optical path imaging. Journal of Atmospheric and Oceanic
Technology, 17(6):761–772, June 2000.
[5] D. Burns, T. Hedblom, and Miller T. Modern pavement marking systems: the
relationship between optics and nighttime visibility. In 18th Biennal
Transportation Research Board Visibility Symposium, 2007.
[6] A. Cohen and H. Fischer. Does the windshield wiper impede a driver’s vision?
International Journal of Vehicle Design, 9(4-5):533–541, 1988.
[7] E. Dumont. Caractérisation, modélisation et simulation des effets visuels du
brouillard pour l'usager de la route. PhD thesis, Université Paris V, Novembre
[8] F. Fournela. Visualisation des projections d’eau sur route mouillée. In
Journées des Sciences de l’ingénieur, Dourdan, France, 2003.
[9] K. Garg and S. Nayar. Vision and rain. International Journal of Computer
Vision, 75(1):3–27, October 2007.
[10] M. Green, M. Allen, B. Abrams, and L. weintraub. Forensic vision with
application to highway safety, third edition. Lawyers & Judges Publishing
Company, 2008.
[11] R. Gunn and G. D. Kinzer. Terminal velocity for water droplet in stagnant
air. Journal of Meteorology, 6:243–248, 1949.
[12] D. Ivey, E. Lehtipuu, and J. Button. Rainfall and visibility the view from
behind the wheel. Technical Report 135-2, College Station, TX: Texas
Transportation Institute, 1975.
[13] T. Kurahashi, Y. Fukatsu, and K. Matsui. Method of evaluating visibility
provided by windshield wipers in rainy conditions. In SAE Technical Paper
Series, number 851636, 1985.
[14] J. Lekner and M. Dorf. Why some things are darker when wet. Applied
Optics, 27(7):1278–1280, 1988.
[15] C. MacAdam. Update on the status of splash and spray suppression
technology for large trucks. Technical report, NHTSA, 2000.
[16] J.S. Marshall and W.M.K. Palmer. The distribution of raindrops with sizes.
Journal of Meteorology, 5:165–166, 1948.
[17] R. Morris, J. Mounce, J. Button, and N. Walton. Field study of driver visual
performance during rainfall. Technical Report DOT-HS-5-01172, College Station,
TX: Texas Transportation Institute, 1977.
[18] R. Morris, J. Mounce, J. Button, and N. Walton. Visual performance of
drivers during rainfall. Transportation Research Record, 628:19–25, 1977.
[19] F. Nedvidek, C. Schneider, and E. Kucerovsky, Brannen. Near-infrared
extinction in rain measured using a single detector system. Journal of
atmospheric and oceanic technology, 3:391–399, 1986.
[20] J. Nicholls. Review of UK porous asphalt trials. Technical Report 264,
TRL, 1997.
[21] D. B. Rensch and R. K. Long. Comparative studies of extinction and
backscattering by aerosols, fog, and rain at 10.6 µm and 0.63µm. Applied Optics,
9(7):1563–1573, July 1970.
[22] R. Rogers, M. Lamoureux, L. Bissonnette, and R. Peters. Quantitative
interpretation of laser ceilometer intensity profiles. Journal of atmospheric and
oceanic technology, 14:396–411, 1997.
[23] M. Stoudt and K. Verdam. Retroreflection from sperical glass beads in
highway pavement markings 1: Specular reflection. Applied Optics, 17(12):1855–
1858, 1998.
[24] C. Stow, G. Bradley, K. Paulson, and L. Couper. The simultaneous
measurement of rainfall intensity, drop-size distribution, and the scattering of
visible light. Journal of Applied Meteorology, 30:1422–1435, 1991.
[25] S. Twomey, C. Bohren, and J. Mergenthaler. Reflectance and albedo
differences between wet and dry surfaces. Applied Optics, 25(3):431–435, 1986.
[26] C. W. Ulbrich and D. Atlas. Extinction of visible and infrared radiation in
rain: Comparison of theory and experiment. Journal of atmospheric and oceanic
technology, 2:331–339, 1985.
[27] F. J. R. Van Mook. Driving rain on building envelopes. Technical report,
Faculty of Architecture, Planning and Building of the Eindhoven University of
Technology, 2002.
[28] H. Vasseur and C. J. Gibbins. Prediction of apparent extinction for optical
transmission through rain. Applied Optics, 35(36):7144–7150, December 1996.
[29] K. Verdam and M. Stoudt. Retroreflection from sperical glass beads in
highway pavement markings 2: Diffuse reflection (a first approximation
calculation). Applied Optics, 17(12):1859–1869, 1998.
[30] H. Zwahlen. Driver eye scanning behavior in rain and during an
unexpected windshield wiper failure. Zeitschrift für Verkehrssicherheit, 26:148–
155, 1980.
... Aside from variation in bead performance, the principal drawback to glass beads is their decreased visibility via retroreflection when water from rain covers the surface of the bead [11]. The reason retroreflection is reduced is twofold. ...
... First, a fraction of the light is lost due to specular reflection of the light from the water surface. Second, retroflection is reduced from the glass beads due to the similarity of indices of refraction of both water and glass [11]. However, recent advancements in the design of the beads have improved retroreflection while ...
... The change in refractive index from water to glass is not as much as the change from air to glass, which diminishes the refraction and overall retroreflection of the bead (Fig. 1). 3 M's new bead has a higher refractive index which makes up for moisture being present on the surface of the sphere. This and other improvements still do not account for the light reflected away from incidence on the flat surface of standing water [11]. A major downside to 3 M's beads is that it only functions when wet, meaning there is no retroreflectance under dry conditions [13]. ...
Roadway marking visibility has made substantial improvements since the beginning of the 20th century. The most common solution for increasing the visibility of road markings is the use of retroreflective beads, which suffer from decreased performance under wet conditions. To remedy this, alternative road marking visibility techniques such as luminescent paint have been investigated. A promising new technology involves the use of phosphors such as strontium aluminate doped with europium and dysprosium ions (SrAl2O4:Eu²⁺,Dy³⁺). This technology shows potential to be seen throughout the night. The two hurdles blocking the path for the use of phosphors in road marking paint are the need to stabilize the particles in the paint matrix, and the tendency of the phosphor to hydrolyze. In this review, an explanation of the science behind stabilization is set forth alongside several methods that show potential for mitigating hydrolysis.
... From the perspective of automobiles, in the case of studies related to night visibility in bad weather conditions, there are studies on subjective and objective evaluation methods and research cases for improving visibility [9][10][11][12][13][14][15][16][17]. In the case of studies on visibility evaluation methods, there is a study on the performance of LED streetlights with different color temperatures [9], and there is a case where the visual performance with different CCTs (correlated color temperatures) in mesopic conditions is evaluated [10]. ...
... These two studies were limited to the contrast between the lane markings and the road surface, not the study on the contrast between the pedestrian and the road surface in bad weather conditions. In adverse weather conditions, especially in fog and rain conditions, the backscattering effect can cause deterioration of driver visibility [15]. In this regard, interesting study related to objective visibility evaluation is a method of estimating visibility in fog conditions at night by detecting the halo of streetlamps and the backscattering veil of headlights using a camera and image processing technique [16]. ...
... Of course, it is also necessary to study backscattering in dense fog conditions. In this regard, as mentioned in the introduction section, several studies have been conducted [15][16][17]. In view of this, it is expected that progress will be ultimately made through sensor-integrated headlamps with respect to issues related to backscattering. ...
Full-text available
This study evaluated a method of applying color temperature convertible headlamps to improve driving safety in adverse weather conditions such as fog and rain during night driving. The concept of color temperature convertible headlamps is to improve the driver’s visibility by driving with a color temperature of 6000 K on a clear night and switching to a color temperature of 3000 K with better light transmittance at night in adverse weather. Through this study, a method for evaluating the night visibility related to such color temperature convertible headlamps under bad weather at night was suggested. To this end, a method of using a facility that can implement weather conditions such as fog and rain was proposed, and evaluation conditions according to the climatic conditions and the distance of pedestrian targets were set and actual tests were conducted.
... Researchers concluded that there was large variation in performance of the paint materials owing to uncontrollable factors such as type of binder, surface conditions and large temperature variations. Another drawback of glass bead reflector is their reduced visibility via retroreflection during rain and standing water on the road surface (Smadi et al., 2014;Hautière et al., 2009). ...
Full-text available
Road infrastructure has witnessed incremental changes in the past as compared to the immense development witnessed by the vehicle’s safety technology. Bott’s dots and other reflector devices are extensively used on the road infrastructure for lane separation and for improving edge detection. These devices come in a large variety of shapes and sizes, however, all of them fall under the category of retroreflectivity since they depend on vehicle lights to provide reflection. Glow-in-the-dark (GiD) material has the benefit that it can store energy during the presence of light and can emit the stored energy in the form of visible light in the absence of an external light source. In this regard, the presented research work details the development and testing of GiD concrete based markers that can be used for lane separation and edge detection. The benefit of the presented innovation is that GiD concrete based markers can be used for visible light instead of retroreflectivity in addition to acting as a driver alertness tool. The durability performance of the presented innovative GiD based raised pavement markers has been presented along with cost comparison to traditional Bott’s dot. In addition, the presented prototype can be adopted for various architectural and esthetical applications in buildings, parks, walkways and bicycle lanes etc.
... Poor recognition of road markings in rainy conditions remains a meaningful challenge for both humans and MV. The phenomenon responsible for this, light scattering and reflection by water particles [47], which is causing significant increase in noise as compared to usable information, simultaneouosly affects human vision and remains a technological obstacle for all MV equipment. This results in significant increase in noise as compared to usable data. ...
The same features of road markings – retroreflectivity and daytime visibility – are the key parameters for their recognition by both human drivers and for machine vision (MV) utilised by the emerging technology of automated vehicles. For the purpose of side-by-side performance assessment of various road marking types, 8 materials, differing in colour and retroreflectivity, were tested under laboratory conditions for visibility by LiDAR and by camera under different weather conditions. Visibility was evaluated under various intensities of rain and fog; simulated effects of glare from oncoming vehicle were also tested. The response of MV equipment depended on (1) the equipment itself, (2) retroreflectivity of road marking, (3) their structure, (3) their colour, and (4) the utilised glass beads. Overall, the highest MV intensities were measured with structured cold plastic reflectorised with ‘premium’ glass beads (refractive index 1.6–1.7) and with white road marking tape. Poorest outcome gave greyish paint imitating severely worn markings and orange paint. Under the conditions of wetness and fog, use of the ‘premium’ glass beads resulted in significantly improved camera contrast ratio, but there was no such correlation with LiDAR intensity. On average, introduction of moisture lowered the measured contrast ratio by 80 % (range 69 %–86 %) and LiDAR response intensity by 84 % (range 72 %–96 %). Results from this case study can be used for development of road marking materials with improved recognition by MV.
... Mortimer and Cameron noted that color-coding lamps by function tended to improve the detection of other signals as (Cameron, 1995;Mortimer, 1995). In addition, it is reported that weather conditions such as fog and heavy (Cavallo et al., 2001;Hautière et al., 2009;Mori et al., 2006), illumination conditions such as daytime and night-time (Khumalo, 2014;Plainis et al., 2006), and observation conditions such as viewing angle (Sivak et al., 2000;Tobitani et al., 2013), affect several visibility indices. These include subjective evaluation, RT, and distance estimation. ...
Full-text available
In recent years, there has been concern regarding the frequent occurrence of nighttime collisions. This research aims to clarify how differences in tail lamp design affect their visibility to other drivers, with a view to develop a design that will make tail lamps more conspicuous. By focusing on visibility in human face recognition abilities, we conducted a subjective evaluation experiment and visual search task using rear-shots of vehicles. The experimental results revealed that a human’s impression of a rear-shot of a vehicle is structurally similar to their impression of a face, and that the tail lamp design affects reaction time. Moreover, electroencephalogram (EEG) measurements verified the validity of the results from a neuroscience perspective. These findings may be used for developing tail lamp designs that are more striking so as to be more noticeable to fellow drivers.
... Then, the presence of droplets reduces the solar energy input and the system's energy efficiency due to reflection and/or absorption of the incident radiation by the droplets [9] . Fogging and rain also hinder the visibility through vehicle windshields [10] and architectural glass windows [11] . Furthermore, light scattering by condensed water droplets can cause severe image distor-tion on camera lenses used for navigation or surveillance [12] as well as on analytical and medical optical instruments [13] . ...
This study investigates systematically the effects of nonabsorbing droplets on the bidirectional transmittance of transparent windows. The Monte Carlo ray-tracing method was used to predict the bidirectional transmittance of transparent windows supporting monodisperse or polydisperse and randomly distributed droplets on their front side or backside. In both cases, photons that did not interact with the droplets were transmitted in the same direction as the incident direction. In addition, the bidirectional transmittance was found to be independent of the diameter and size distribution of the nonabsorbing droplets. It also featured cutoff transmission angles for certain ranges of incident and contact angles for either external or backside droplets. Analytical predictions of these cutoff angles were developed and were in excellent agreement with results from numerical simulations. Moreover, the bidirectional transmittance of the windows with either front or back side droplets increased with increasing projected surface area coverage at transmission angles other than the incident angle. The hemispherical distributions of the bidirectional transmittance showed that the photons scattered into other transmission azimuthal angle were concentrated in the angular region bounded by the cutoff angles for either external droplets or backside droplets.
... Torrential rain events have recently become more common in Japan [1,2], along with increasing traffic problems caused by unusual weather [3]. Heavy rain raises driving risk [4][5][6] due to lowered visibility [7,8] and road friction [9]. While localized short-term heavy rain is difficult to predict [10], its narrow coverage makes it possible to avoid through altered travel routes. ...
Full-text available
Localized torrential rainfall events and related traffic problems are increasing in Japan, suggesting the need for a navigation-alert system to help drivers avoid such risks. Based on ongoing developments of weather radar systems for early detection of localized torrential rain and a cross-data collaboration platform for traffic optimization, in this study we tested the application of a route-guidance system that can help drivers avoid heavy rainfall. Participants were given equivalent levels of pre-training un the early detection of rainfall and the relationship between rainfall and accidents, then allowed to test a driving simulator set up with four alert methods, three route options, and four levels of possible risk avoidance. Using this system, the heavy rain avoidance rate was 85.63%, suggesting that such a system would be socially acceptable and useful, though further research is needed to refine the specific approach.
... Therefore, these factors could also influence the subjective driving evaluation. In terms of situational factors, (oncoming) traffic (Schiebl, 2008;Teh et al., 2014) and rain (Ashley, Strader, Dziubla, & Haberlie, 2015;Hautière, Dumont, Brémond, & Ledoux, 2009) appear to be relevant. However, even though rain is the weather condition having the greatest impact on the accident risk, weather in general plays a minor role in terms of accident causes (Dingus et al., 2006). ...
Lately, the development and implementation of automated driving moved to the center of interest in the automotive industry. In this context, one of the central issues – the configuration of adequate trajectories – is mainly tackled using a technical approach. However, it appears that a technically ideal driving performance does not necessarily coincide with the drivers’ subjective preferences. This study strives to determine thresholds of a subjectively accepted driving performance regarding lateral vehicle control. A second objective is to analyze the influence of selected personal and situational factors on these thresholds. An empirical online survey with 161 participants rating video sequences of driving performances was conducted. The video sequences differed not only with regard to the lateral offset of the ego-vehicle but also concerning the weather (sun/rain) and traffic conditions (existence/driving behavior of oncoming traffic). Additionally, the participants’ driving experience and sensation seeking were considered in the data evaluation. To analyze the data, binary logistic regression analyses were calculated. They revealed that the subjective evaluation of driving performances varies primarily depending on the lateral offset of both the ego-vehicle and the oncoming traffic. The results indicate that regarding the lateral offset certain thresholds of subjectively accepted driving performances do exist. Regarding the development of automated driving systems, two issues need to be considered in order to ultimately guarantee user acceptance. First, the subjective thresholds need to be integrated into the systems’ trajectory planning. Second, the oncoming traffic’s driving behavior has to be considered.
Full-text available
Road lane markings play an essential role in maintaining traffic order and improving traffic safety and efficiency. Active luminous lane markings have emerged with advances in technology recently. However, it is still not completely clear what impact their application will have on drivers. This paper aimed to study the effectiveness of active luminous lane markings on highways at night. A driving simulation experiment was carried out based on advanced driving simulators at Tongji University. The driving simulation experiment involved 31 participants and 9 simulation scenes with 6 different types of lane markings models and the same 2-way highway segment, which was 5300-m long with four 3.75-m wide driving lanes. The study participants drove through the simulated highway while the vehicle operation data and the driver’s eyes changing data were continuously captured. Overall, the pupil area change rate, steering wheel speed, brake pedal force, gas pedal, lane departure, and operating speed indicators were selected to evaluate the effectiveness of the active luminous lane markings. The results are shown as follows: (1) the active luminous lane markings have excellent visual recognition performance at night. Compared with the passive luminous lane markings, the active luminous markings can reduce the mental and physical loads of drivers, increase the early braking distance significantly, improve the lane-keeping ability and smooth the operating speed; (2) for the specific parameter settings of the active luminous lane markings at night, the yellow lane markings are better than the white ones, the point-line-type lane markings are superior to the conventional-type ones, and the blinking frequency is reasonable to set, at a moderate level, as 40 times per min. The results suggest that there are positive effects of active luminous lane markings on the promotion of highway traffic safety and efficiency at night, providing theoretical support for the popularization and application of active luminous road lane markings.
Full-text available
Individuelle Mobilität ist ein zentrales Thema menschlicher Gesellschaften. In diesem Kontext entwickelte sich der PKW zum primären Fortbewegungsmittel. Die Ausführung der Fahraufgabe stellt in diesem für Fahrerinnen und Fahrer bereits eine hohe Belastung dar. Trotzdem bearbeiten sie oft zusätzliche fahrfremde Tätigkeiten (FFT). Aufgrund der begrenzten menschlichen Kognitionsressourcen kann diese parallele Bearbeitung mehrerer Aufgaben zu Fahrerablenkung führen. Allerdings sind Menschen relativ zu den gefahrenen Kilometern selten in schwere Unfälle verwickelt. Dies legt nahe, dass sie Fähigkeiten zur Unfallvermeidung besitzen. Hierzu konnte Forschung im Kontext des nicht-automatisierten Fahrens zeigen, dass Fahrerinnen und Fahrer ausgehend von ihrem Situationsbewusstsein in Erwartung einer kritischen Fahrsituation bzw. Fahrleistung proaktiv ihre (kognitiven) Ressourcen regulieren und von FFT auf die Fahraufgabe verschieben. Verschiedene theoretische Modelle beschäftigten sich mit diesem regulativen Fahrerverhalten. Aufbauend auf diesen stammt ein ganzheitliches Arbeitsmodell von Schwalm, Voß und Ladwig (2015; Voß & Schwalm, 2015), welches das regulative Fahrerverhalten als funktionale Verhaltensanpassungen konzeptualisiert. Während Fahrerinnen und Fahrer im nicht-automatisierten Fahren für die sichere Ausführung und Überwachung der Fahraufgabe zuständig sind, ist eine solche dauerhafte Involvierung im automatisierten Fahren je nach Automationsgrad nicht mehr erforderlich. Fahrerinnen und Fahrer können sich mit FFT beschäftigten (ab SAE Level 3) und das Situationsbewusstsein der Fahrerinnen und Fahrer sinkt ab. Dennoch dürfen sie jederzeit in die automatisierte Fahrzeugführung eingreifen bzw. werden bis zu einem bestimmten Automationsgrad sogar als Rückfallebene benötigt (bis SAE Level 3). Aus dieser Kombination eines reduzierten Situationsbewusstseins und den möglichen Fahrereingriffen im automatisierten Fahren ergibt sich die Frage, wie Fahrerinnen und Fahrer es schaffen, in solchen Situationen eine sichere Fahrleistung zu gewährleisten und ob sie zu diesem Zweck auch hier auf die funktionalen Verhaltensanpassungen zurückgreifen können. Die vorliegende Dissertation nimmt sich dieser Thematik an. Es wird die Zielsetzung (a) der theoriebasierten und empirischen Ausarbeitung ausgewählter Komponenten des Arbeitsmodells der funktionalen Verhaltensanpassungen von Schwalm et al. (2015; Voß & Schwalm, 2015) als theoretischer Referenzrahmen der Dissertation sowie (b) der spezifischen Untersuchung der Verfügbarkeit und Ausprägung derselben im Rahmen des automatisierten Fahrens verfolgt. Hierzu wurde das Arbeitsmodell zunächst theoriebasiert detailliert. Es wurde herausgearbeitet, dass die funktionalen Verhaltensanpassungen im Mehrfachaufgabenkontext vor allem in Abhängigkeit der Situationswahrnehmung sowie der subjektiven Fahrleistungsbewertung auftreten. Anschließend wurden Annahmen zur Funktionsweise der funktionalen Verhaltensanpassungen im automatisierten Fahren getroffen. Es wurde postuliert, dass Fahrerinnen und Fahrer im Falle von Übernahmen proaktiv die Bearbeitung von FFT zur Freigabe kognitiver Ressourcen reduzieren, welche anschließend für das sichere Lösen der Fahraufgabe genutzt werden. Diese Annahmen wurden anschließend empirisch geprüft. In einer Fahrsimulationsstudie (Studie 1) wurden die funktionalen Verhaltensanpassungen in Abhängigkeit der Situationswahrnehmung in einer sich verändernden Fahrsituation (Übernahmesituation vom automatisierten zum nicht-automatisierten Fahren) untersucht. Es zeigte sich, dass Fahrerinnen und Fahrer gemäß den theoretischen Annahmen vor einer Übernahme proaktiv die FFT reduzierten, hierüber kognitive Ressourcen freigaben und somit eine sichere Übernahme ermöglichten. Die folgenden Studien untersuchten die Idee, dass solche funktionalen Verhaltensanpassungen ebenfalls bei Abweichungen von subjektiv akzeptierten Trajektorien auftreten können. Zunächst wurde das Konstrukt einer subjektiv angemessen empfundenen Fahrleistung diskriminanz- und faktorenanalytisch geprüft (Studie 2) und Schwellenwerte subjektiv akzeptierter Fahrleistungen hinsichtlich des Lateralversatzes in Abhängigkeit diverser Personen- und Situationsfaktoren bestimmt (Studien 3 und 4). Anschließend wurde in den Studien 5 und 6 die Handlungsrelevanz der Fahrleistungsschwellen im Mehrfachaufgabenkontext des automatisierten Fahrens im Fahrsimulator und unter Realbedingungen auf einer Teststrecke untersucht. Bei Überschreitungen der Schwellenwerte einer subjektiv angemessen empfundenen Fahrleistung zeigten sich dort nicht nur Komforteinbußen, sondern auch die erwarteten funktionalen Verhaltensanpassungen. Im Anschluss an eine proaktive Reduktion der FFT griffen Fahrerinnen und Fahrer vermehrt in die automatisierte Fahrzeugführung ein. Die Eingriffe waren teilweise allerdings nicht optimal bzw. sogar sicherheitskritisch. Die Erkenntnisse der sechs empirischen Studien erlaubten abschließend Schlussfolgerungen zu der Verfügbarkeit funktionaler Verhaltensanpassungen im automatisierten Fahren. Weiterhin wurde zukünftiger Forschungsbedarf, wie die fortführende Modellvalidierung oder die konkrete Gestaltung automatisierter Systeme zur Unterstützung der funktionalen Verhaltensanpassungen, identifiziert. Individual mobility is a central theme of human societies. In this context, the car has become the primary means of transport in which the execution of the driving task already constitutes a high task load for drivers. Nevertheless, they often work on non-driving related tasks. Due to the limited human cognitive resources, this parallel processing of multiple tasks can lead to driver distraction. Yet, people are rarely involved in serious accidents relative to the kilometres driven. Drivers thus seem to have abilities for the avoidance of accidents. Research in the context of non-automated driving can support this claim. Studies could show that drivers - based on their situation awareness - proactively regulate their (cognitive) resources and shift them from the non-driving related tasks to the driving task in case they expect a critical driving situation or driving performance. Various theoretical models dealt with this regulative driver behaviour. Based on these, a holistic working model originates from Schwalm, Voß and Ladwig (2015; Voß & Schwalm, 2015). It conceptualises this regulative driver behaviour as functional behavioural adaptations. While in non-automated driving drivers are responsible for the safe execution and monitoring of the driving task, in automated driving such a permanent involvement in the driving task is no longer necessary, depending on the degree of automation. Drivers can work on non-driving related tasks (from SAE Level 3 on) and drivers’ situation awareness decreases. However, they are allowed to intervene into the automated vehicle guidance at any time or are even required as fall-back option up to a certain degree of automation (until SAE Level 3). This combination of a reduced situation awareness and possible driver interventions in automated driving raises the question of how drivers guarantee a safe driving performance in such situations and whether they can make use of functional behavioural adaptations. This doctoral thesis deals with this topic. The objectives are (a) the theory-based and empirical elaboration of selected components of the working model regarding functional behavioural adaptations from Schwalm et al. (2015; Voß & Schwalm, 2015) as theoretical frame of reference of this doctoral thesis, and (b) the specific investigation of the availability and characteristics of these in the context of automated driving. For this purpose, the working model was detailed theory-based. It was highlighted that the functional behavioural adaptations in the multitasking context particularly occur depending on the perception of a situation and depending on the subjective evaluation of a driving performance. Following, assumptions were made on how functional behavioural adaptations work in the context of automated driving. It was postulated that in case of takeovers, driver proactively reduce activity in non-driving related tasks to release cognitive resources, which subsequently are used for the safe execution of the driving task. These assumptions were empirically assessed. In a driving simulation study (study 1), the functional behavioural adaptations were examined as a function of the perception of a changing driving situation (takeover from automated to non-automated driving). According to the theoretical assumptions, drivers proactively reduced the processing of a non-driving related task before a takeover, released cognitive resources, and thus enabled a safe takeover. The following studies investigated the idea that such functional behavioural adaptations can also occur in case of deviations from a subjectively accepted trajectory. Initially, the construct of a subjectively accepted driving performance was examined by means of discriminant function and factor analysis (study 2). Thresholds of such a subjectively accepted driving performance regarding the lateral offset as a function of various personal and situational factors were examined (studies 3 and 4). Subsequently, studies 5 and 6 investigated the relevance for action of the thresholds in the multitasking context of automated driving in a simulator and under real conditions on a text track. In case the thresholds were exceeded, not only comfort losses but also the expected functional behavioural adaptations occurred. After a proactive reduction of the non-driving related task, drivers often intervened in the automated vehicle guidance. Some of the interventions, however, were not optimal or even safety critical. These insights of the six empirical studies allowed for conclusions regarding the availability of functional behavioural adaptations in the context of automated driving. Furthermore, future research needs were identified, for example a continued working model validation or the design of automated systems which support the functional behavioural adaptations.
Full-text available
Field measurements were made of the attenuation of a low-power. He-Ne laser beam over a 272-m path. Concurrently, high-resolution (10 s) measurements of rainfall intensity were obtained at several points along the path and drop-size distributions derived from 100-s samples updated every 10-s. A double Gaussian is shown to be a reasonable approximation to the small-angle scattering phase function expected in Marshall-Palmer-type rain, but this phase function is only partially successful in reconciling rainfall intensities and beam attenuation. Considerable departures from a Marshall-Palmer size distribution were found on occasion but wore not necessarily associated with this lack of reconciliation. It is conjectured that inconsistencies between theory and experiment are due to the spatial inhomogeneity of rain over distances as short as 50 m.
The factors influencing wet weather accidents are complex, a fact sometimes ignored during the current popularity of blaming low tire pavement friction for most wet weather accident problems. A factor of significance is visibility, as influenced not only by rainfall intensity but to a great extent by traffic speed. This report presents a number of direct visibility observations and develops a framework useful in interpreting these data to determine the influence of reduced visibility on the operation of motor vehicles. Information from the literature shows the low probability of high intensity rainfalls. Conclusions concerning the hazard of passing maneuvers during rainfall of 1 inch per hour or more, and the need to reduce speed under wet weather conditions, are also presented.
This paper reports an investigation of the effect of rain on the visual performance of drivers. The degradation of static visual acuity in terms of visual angle, detection probability, and legibility as a function of rain intensity was determined by experiments that used a rainfall simulator that produced artificial rain.
Does the movement of windshield wipers impede a driver's vision? To study this question drivers' eye movements were recorded with windshield wipers off and on. It was assumed that impeded vision would be manifested by a differing pattern of eye movements in the two conditions investigated. The quantitative data for the eye movements were analyzed. This consisted of, for example, the fixation times, the amplitudes of the saccadic movement (as well as its horizontal and vertical components) and the relationship between the position of the eyes and that of the head. The results suggest that operated windshield wipers do not impair vision. However, the relation between the eyes and the head is affected when the windshield wiper is on, which we suppose to be a result of a compensatory process. The results are discussed in terms of focusing attention on the road which is accompanied by binocular phenomena like suppression of one foveal image, when the two images range beyond the Panums area whereby the prevalence of the image focused on facilitates the uninterrupted information flow from the scenery ahead. The maintenance of uninterrupted relevant information flow is, we suppose, facilitated by processes that favour information being targeted at the focus of visual attention.
The equilibrium shape of raindrops has been determined from Laplace's equation using an internal hydrostatic pressure with an external aerodynamic pressure based on measurements for a sphere but adjusted for the effect of distortion. The drop shape was calculated by integration from the upper pole with the initial curvature determined by iteration on the drop volume. The shape was closed at the lower pole by adjusting either the pressure drag or the drop weight to achieve an overall force balance. Model results provide bounds on the axis ratio of raindrops with an uncertainty of about 1% and very good agreement with extensive wind tunnel measurements for moderate to large water drops. The model yields the peculiar asymmetric shape of raindrops: a singly curved surface with a flattened base and a maximum curvature just below the major axis. A close match was found between model shapes and profiles obtained from photos of water drops for diameters up to 5 mm. Coefficients are provided for computing raindrop shape as a cosine series distortion on a sphere. In contrast to earlier models of raindrop shape for the oblate spheroid response to gravity (Green, Beard) or the perturbation response to the aerodynamic pressure for a sphere (Imai, Savic, Pruppacher and Pitter), the present model provides the appropriate large amplitude response to both the hydrostatic and aerodynamic pressures modified for distortion. In addition, the new model can be readily extended to include other pressures such as an electric stress.
Four field studies were conducted in natural rainfall to develop a model for predicting distances at which drivers are able to see other vehicles in the roadway at various time periods following stoppage of a windshield-wiper stroke. The seeing distance prediction was developed as a function of rain intensity, rain accumulation time, and ambient daylight illumination. Two situations were studied; in the first, drivers seated in a stationary vehicle detected moving vehicles, and in the second, moving drivers detected a stationary vehicle. Useful seeing distance models were developed from the field studies. Seeing distances predicted from the models developed from these earlier studies were compared with seeing distances obtained in a subsequent validation field test. Results indicated that average error in the prediction of seeing distances ranges from 9% to 23%.
The performance and operation of an optical device to accurately measure extinction due to rainfall over a 100 m sample path is described. A collimated beam from an infrared light-emitting diode operating at 0.94 mu m is used as a sensing beam. A PIN diode detector receives reference and sample signals alternately in a switch arrangement using a beam splitter and mirrored chopper wheel. Demultiplexing and phase sensitive detection are used to separate and demodulate the sample and reference signals. The experimental results are in agreement with theoretical predictions and theoretical results obtained for rainfall rates up to 90 mm h(-1). Extinction calculations based on the recent theoretical treatment of Ulbrich and Atlas produced a best fit to the experimental results.