Horizontal road markings for human and machine vision

Abstract

Horizontal road markings are inalienable feature on almost all roads because their presence results in significant increase in safety for all road users. By providing delineation, they help drivers in keeping the vehicle in the traffic lane. At night time, retroreflectivity (RL) of road markings, achieved by incorporating glass beads on the marking surface, is perceived by the drivers and its level was reported to be associated with lower accident rate. Indeed, usable life of road markings is measured by their RL. It is shown herein, based on field tests, that for thick-layer structured road marking systems the selection of glass beads has profound effect not only on retroreflectivity, but also on durability. Exemplary financial analysis demonstrates that despite higher one-time expense, savings can be realised in the long-term with the use of premium materials. Moreover, the same high quality and high RL of horizontal road markings that are needed for human drivers are demanded by machine vision algorithms, which provides guidance in the emerging autonomous vehicles technology. Therefore, maintaining of horizontal road markings at high level is necessary-until all of the vehicles would be self-driving, and very possibly even afterwards. Furthermore, because of reasonable price and broad availability, durable horizontal road markings have a potential of being one of the solutions to lower accident occurrences in poorly developed countries.

Full text

Available online at www.sciencedirect.com
ScienceDirect
Transportation Research Procedia 00 (2019) 000–000
www.elsevier.com/locate/procedia
PRE-PRINT.
Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089
2352-1465 © 2020 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Peer-review under responsibility of the scientific committee of the World Conference on Transport Research – WCTR 2019
World Conference on Transport Research – WCTR 2019, Mumbai, 26-30 May 2019
Horizontal road markings for human and machine vision
Tomasz E. Burghardta
*
, Harald Mosböckb, Anton Pashkevichc, Mario Fiolićd
aM. Swarovski Gesellschaft m.b.H.; Industriestraße 10, 3300 Amstetten, Austria
bSwarco AG; Blattenwaldweg 8, 6108 Wattens, Austria
cPolitechnika Krakowska; ul. Warszawska 24, 31-155 Kraków, Poland
dUniversity of Zagreb, Faculty of Traffic and Transport Sciences; Vukelićeva 4, 10000 Zagreb, Croatia
Abstract
Horizontal road markings are inalienable feature on almost all roads because their presence results in significant increase in safety
for all road users. By providing delineation, they help drivers in keeping the vehicle in the traffic lane. At night time, retroreflectivity
(RL) of road markings, achieved by incorporating glass beads on the marking surface, is perceived by the drivers and its level was
reported to be associated with lower accident rate. Indeed, usable life of road markings is measured by their RL. It is shown herein,
based on field tests, that for thick-layer structured road marking systems the selection of glass beads has profound effect not only
on retroreflectivity, but also on durability. Exemplary financial analysis demonstrates that despite higher one-time expense, savings
can be realised in the long-term with the use of premium materials. Moreover, the same high quality and high RL of horizontal road
markings that are needed for human drivers are demanded by machine vision algorithms, which provides guidance in the emerging
autonomous vehicles technology. Therefore, maintaining of horizontal road markings at high level is necessary — until all of the
vehicles would be self-driving, and very possibly even afterwards. Furthermore, because of reasonable price and broad availability,
durable horizontal road markings have a potential of being one of the solutions to lower accident occurrences in poorly developed
countries.
© 2020 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Peer-review under responsibility of the scientific committee of the World Conference on Transport Research – WCTR 2019
Keywords: horizontal road markings; retroreflectivity; road safety; road maintenance financing; autonomous vehicles; machine vision
1. Introduction
Transport of people and goods is a key human activity, necessary for economic development and to maintain the
quality of life. Unfortunately, due to conflicts in traffic and human inattention, accidents occur quite frequently.
* Corresponding author. Tel.: +43 664 8878 4307
E-mail address: tomasz.burghardt@swarco.com

2 Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT.
According to World Health Organization (2016), over 3,400 people die every day in road accidents. Furthermore, the
same report estimates that financial expenses associated with vehicular accidents reach an enormous 3% of World’s
Gross Domestic Product. Therefore, increase in road safety should be of profound importance for all countries and
one of the ways to improve standard of life for all people. The most endangered are unprotected road users in middle-
and low-income countries. It was noted that for those countries, comprising 84% of Earth’s population, there are 53%
of registered vehicles, but that is where 92% of road deaths occur. Road fatality rates per 100,000 citizens are: 24.1 in
Africa, 20.1 in middle-income countries, 8.7 in high-income world, and 5.2 in European Union countries (Toroyan et
al., 2013). Given the enormous social cost and the financial burden caused by the crashes, an inexpensive and effective
solutions for improvement of road safety is necessary, especially in these countries.
Horizontal road markings could be such a solution because of the positive effect on road safety that could be
achieved at low cost. Herein, we are providing the results from field testing of premium quality road marking materials
having high durability, which lowers the long-term financial expenses. Simultaneously, with the recent technological
advances, Autonomous Vehicles (AV) can be considered as an answer to high accidents rates; however, it must be
noted that because AV most of the time rely on Machine Vision (MV) for guiding, they cannot function efficiently on
poorly maintained roads. Thus, a solution to successful MV is improvement of the same horizontal road markings that
are needed for human vision and safety on the road.
2. Horizontal road markings
2.1. Road markings systems
Horizontal road markings are systems, comprising the base layer and the retroreflective layer (Pocock and Rhoades,
1952). The base can be divided into thin-layer, usually applied at thickness below 1 mm, and thick-layer, which can
reach several millimetres. The base layer can be a solventborne or a waterborne paint that dries due to evaporating
solvent, a thermoplastic mass that is applied as a hot melt, a cold plastic that polymerizes on the road surface, or plural-
component materials with components undergoing chemical reaction on the road surface (Babić et al., 2015). The
base layer is a critical component of a road marking system, because it gives the desired colour, forms surface for
retroreflection, and holds glass beads. While field studies have shown that the selection of the base system is
controlling the durability of thin-layer road marking systems (Burghardt et al., 2016), there are no similar systematic
studies related to thick-layer applications reflectorised with different glass beads. Thick-layer road markings are
usually applied as either regularly or stochastically distributed structures. With structured marking, a vibroacoustic
effect, which warns drivers deviating from the traffic lane, can thus be achieved. In addition, the presence of the
structure facilitates water drainage, which improves RL under wet conditions and also shelters some of glass beads
from the action of passing vehicles and snow ploughs. Thick-layer markings because of their expense are most suitable
for roads with high traffic load.
The second inalienable component of horizontal road marking systems are glass beads, which provide
retroreflection and protect the base layer from abrasion. A very important property of glass beads designed for road
markings is their refractive index (RI), which varies from 1.5 for standard materials prepared from recycled glass to
1.9 for high-index materials prepared from a virgin glass melt, for special applications. A unique type of quite recently
developed premium glass beads has RI increased to 1.6-1.7; their performance shall be discussed herein. Those
premium glass beads are prepared from in a proprietary process from a virgin glass melt, using carefully selected raw
materials that furnish improved resistance to scratching. They are meeting Class 1 and Class A normative requirements
set in EN 1423 (European Committee on Standardization, 2012). Microphotograph of standard glass beads, with their
occasional imperfections, is shown in Figure 1 and exceptional roundness and flawless surface of premium glass beads
can be seen in Figure 2.
With properly embedded standard glass beads used to reflectorize white paint one can expect initial RL around 400
mcd/m²/lx, while premium beads can furnish RL about 1,000 mcd/m²/lx. Higher RL, reaching even 2,000 mcd/m²/lx,
can be readily initially obtained with high-index glass beads; however, due to their low resistance to scratching and
very high price, they are a niche application.

Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT. 3
Figure 1. Standard glass beads, fraction 630-700 µm.
Figure 2. Premium glass beads, fraction 630-700 µm.
2.2. Retroreflectivity as the main feature of horizontal road markings
Horizontal road markings were reported as one of the most effective measures to improve road safety. Miller (1992)
has calculated that the financial expense of installation and proper maintenance of road delineation was on average
sixty times lower than the costs caused by chaotic traffic and accidents. Such benefit is possible, because driving is a
task based on visual input, so there is a necessity of providing information to drivers about position of their vehicles
and horizontal delineation serves that purpose. Steyvers and de Waard (2000) demonstrated that the presence of edge
lines markings helps drivers maintain the position of vehicle in the centre of the lane. The effect was found by Calvi
(2015) to be particularly prominent at curves.
Horizontal road markings are especially important during night time driving on unlit roads, because that is when
the relative number of accidents and their severity are significantly increased despite much lower traffic loads (Plainis
et al., 2006). At night, when the number and quality of visual cues are meaningfully limited, the guiding role of
horizontal delineation becomes more prominent. However, the markings must be reflectorised to be visible in the
vehicles’ headlights, which requires the use of glass beads. Retroreflectivity occurs when the light is reflected back
toward the driver and its level depends on the properties of the road marking system, particularly the glass beads type,
quality, and embedment, and the colour of the markings. Zwahlen and Schnell (1999) reported that during night time
driving, RL becomes a natural focus point for all drivers. RL is measured according to standard norms and reported in
millicandela per square metre per lux (mcd/m²/lx) and its level is used to determine the usability of road markings.
Therefore, one cannot discuss horizontal road markings without their RL, except for very limited special applications.
Only recently there were published results from a complex multi-year statistical analysis correlating the number of
accidents with the level of RL. Carlson and co-workers (2013) found that up to 23% reduction of single-vehicle non-
weather-related accidents between intersections could be realised with increasing RL by 100 mcd/m²/lx. Subsequently,
the calculation methodology was confirmed (Carlson et al., 2015). Bektas et al. (2015), using slightly different
evaluation protocol, also established the relationship, but not at all roads and not as strong. Nonetheless, reports
doubting the correlation between RL and road safety should be noted as well (Hauer, 2019). However, there are no
articles addressing the possible safety benefits of very high levels of RL, which could be achieved with the novel
technology that is discussed below.
3. Autonomous driving and machine vision
Approximately 90% of accidents are currently attributed to a human error (Dingus et al., 2016). Therefore, it is
generally accepted that replacement of human drivers with infallible computers equipped with sensors would eliminate
those crashes. Whereas there is a technological capability and policy push for AV, it sometimes appears to be forgotten

4 Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT.
they cannot work properly without appropriately maintained infrastructure, because they rely mostly on MV to
determine the location within the travel lane.
3.1. Machine vision and horizontal road markings
With the plethora of literature related to AV and MV, notable and quite surprising is lack of visible collaboration
between the scientists developing these new advanced technologies and the researchers working on horizontal road
markings. Yet, MV is mostly dependent on the quality, clarity, and RL of horizontal road markings and some of the
major obstacles in broad implementation of AV are related to inadequate recognition of travel path.
There are only a few studies directly relating those two factors. About a decade ago, performance of lane departure
warning systems was reported to improve with the increase of line RL (Hadi et al., 2007). It was later recognised that
RL of road markings is needed for all vision-based MV systems for proper night time guidance (Hadi and Sinha, 2011).
A patent was issued to inventors who disclosed a pathway determination based on RL (Stroila et al., 2014). Matowicki
and co-workers (2016) reported a malfunction of a MV equipment on a road with markings having RL of only 50
mcd/m²/lx. Davies (2017) presented the results of laboratory testing of MV with various horizontal markings:
detection was reported to depend on line width (broader lines gave better results, even at lower RL), RL (higher RL
was easier to detect), and colour (yellow was more difficult to recognise than white); rather poor results were obtained
during laboratory-induced rain. Carlson and Poorsartep (2017) after a field experiment reported on the noted
deficiencies, which were noticed when using MV for road marking systems presently applied in North America. For
MV, the main identified issues were lack of line detection in case of poor contrast and incorrect line assignment in
case of poorly maintained roads and in strong sunshine (Carlson, 2017). All of these deficiencies do occur because
MV relies on successful pattern recognition; recent advances in that field were summarised by Narote and colleagues
(2018).
There are several critical path recognition issues, which are handled quite well by humans, but must be solved for
MV. Various issues addressed below, associated with the quality of markings, are relatively easy to address with the
current technology and results furnished herein can be used as the desired solution.
• Absence of horizontal markings in non-residential areas, which is really frequent on rural roads with very low
traffic or on narrow ones.
• Obstruction of the markings by snow, ice, dirt, other vehicles, vegetation, etc.
• Lack of standard marking type, which demands complex programming of the MV software for proper
classification.
• Poor condition of road surface, where cracks, potholes, and patches may lead to a misperception by the MV.
• Various marking quality, which seldom confuse people, but can confuse MV and cause inappropriate reading.
• Weather-related insufficient visibility of markings, in either inclement weather (snow fall, rain downpour, dense
fog) and in excellent sunny weather (due to insufficient or excessive contrast and glare).
• The presence of ‘phantom markings’, caused either by removal of temporary markings, modification of the
driving paths, or by longitudinal cracks or seal lines on the roadway surface.
• Proper handling of temporary markings, with all of their meanings and frequent lack of clarity. While there is a
multitude of reports and patents related to AV and MV, there is almost a complete absence of literature related to
reading of temporary markings by MV.
3.2. Horizontal road markings for human and machine vision
We are listing below some known issues associated with reading of horizontal road markings by MV and associate
them with the known needs for human vision. It is obvious that solving an inefficiency for a human driver
simultaneously solves it for MV. The current MV technology, despite access to a broad spectrum of sensors, still does
not match human vision, because computers cannot process ambiguous information as correctly as humans. The
premium structured road marking systems that are described herein – clear and highly retroreflective – are meeting
vast majority of these demands. These issues are listed in Table 1.

Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT. 5
Table 1. Horizontal road markings and machine vision issues.
Issue
Description
References
Retroreflectivity
Current advanced MV require over 50 mcd/m²/lx
Carlson, 2017
Minimum RL for human drivers should be above 68 mcd/m²/lx
Parker and Meja, 2003; Burns et
al., 2006
Minimum RL for older drivers should be above 150 mcd/m²/lx
Parker and Meja, 2003
RL>150 mcd/m²/lx recommended for all colours at all times for comfortable driving
Gibbons et al., 2012
Drivers notice and appreciate RL>500 mcd/m²/lx
Burghardt et al., 2017;
Pashkevich et al., 2017
Field experiment demonstrated increased comfort of drivers at roads with high RL, with
particular benefit for elderly drivers
Diamandouros and Gatscha,
2016
Laboratory experiment proved increased comfort of drivers with enhanced RL
Horberry et al., 2006
European Road Federation proposed that a minimum of 150 mcd/m²/lx should be
maintained at all lines at all times
European Road Federation,
2015
Line width
Wider markings facilitate their detection by MV and limit the number of false detections
Davies, 2017; Carlson, 2017
Statistical analysis shown that up to 38% accident reduction could occur with lines 15
cm wide (as compared to lines 10 cm)
Park et al., 2012
Lines 15.2 cm wide are now required in California for all of the lines on state roads
State of California, 2017
Lines 15 cm wide are recommended by European Road Federation
European Road Federation,
2015
Contrast ratio
Contrast ratio between the marking and the pavement should be above 2:1, with better
results achieved with MV at 3:1 contrast ratio
Carlson, 2017
Glare is severely impairing driving ability
Theeuves et al., 2002
Glare contributes to increase of crashes at intersections
Hagita et al., 2011; Mitra, 2014;
Sun et al., 2018
Sun glare contributes to increase of collisions at intersections due to signal violations
and mid-block collisions due to improper turning or lane changes
Sun et al., 2018
Glare might cause effects similar to driving in fog, when, due to a perceptual quirk,
drivers think they are driving far more slowly than they actually are, and therefore
increase their speed
Snowden et al., 1998
Clarity
Sharp edges of markings would be advantageous for detection and would eliminate
ambiguities caused by seal marks
Carlson, 2017
Clarity of horizontal road markings was reported to be preferred by human drivers and
through association with aesthetics having influence on perception of road safety and
thus the safety itself
Żakowska, 1997
Unification
Unification of markings across various countries is necessary for reliable MV and
universality of AV
Carlson, 2017
Difficulty of comprehension of vertical road signs by foreign drivers is well-established
and the same occasionally may apply to horizontal signage, particularly for temporary
marking
Shinar et al., 2003
Unification of road signage across countries was proposed as a method of lowering the
number of accidents and fatalities by up to 5,000 annually on Trans-European Road
Network
Räsänen and Horberry, 2006
Raised pavement markings (‘Bott’s Dots’) cannot be reliably classified by MV under all
conditions
Carlson, 2017
‘Bott’s Dots’ are being discontinued on state roads in California
State in California, 2017

6 Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT.
4. Field evaluation of road markings
4.1. Durability of horizontal road markings
Horizontal road marking systems for human and machine vision must meet the requirements of clarity and
retroreflectivity and must be durable to lower environmental burden. Simultaneously, to be a solution for less-
developed countries and be socially accepted in the developed states, they must be reasonably inexpensive. Herein are
presented the results from field testing of such systems, done in three countries. Reflectorisation with three types of
glass beads: standard, 30% premium (mixed with 70% standard), and 100% premium permitted for the assessment of
the influence of glass beads on thick-layers structured cold plastic road markings. Nonetheless, it must be emphasised
that at the three test fields were used base layer materials from different manufacturers, application was done by
different crews, and weather conditions were different, which all could have profound influenced the outcome.
It must be understood that evaluation of road marking systems is difficult due to high level of uncertainty caused
by numerous factors affecting their performance. While we are demonstrating that the selection of glass beads can be
the controlling parameter, it must be acknowledged that the climatic conditions, application quality, and road geometry
can be equally important. Several researchers were struggling with finding correlation parameters for road markings
durability, but the appropriate model comprising majority of factors is still not developed in spite of some progress
(Migletz et al., 2001; Zhang and Wu, 2006; Sitzabee et al., 2009; Hummer at al., 2011; Wang et al., 2016; Babić et
al., 2019). As an example of critical influence of climatic conditions, we can provide the instance of thin-layer
application of the same road marking systems in Croatia and in Poland: whereas in Croatia a three-year durability was
achieved, in Poland the marking was destroyed during winter by snow ploughs (Burghardt et al., 2019).
4.2. Test fields and materials
Cold plastic structured thick-layer systems were applied on major dual carriageway roads in Switzerland, in
Croatia, and in Poland. Application was done by local companies, using their standard procedures and equipment. The
base layer materials, delivered by the applicators, were reflectorised with standard, premium, and premium mixed
glass beads. The tested premium glass beads were SOLIDPLUS (M. Swarovski GmbH; Amstetten, Austria),
characterised by exceptional roundness, improved resistance to scratching, and RI slightly increased to augment RL,
but still remaining within Class A according to the norm EN 1423:2012. Periodic measurements of RL were done by
independent parties using a dynamic method, with a retroreflectometer installed on a vehicle, which collects data
during normal driving every few milliseconds and then provides average results for 50-100 m stretches.
In Table 2 is provided the basic information about the test stretches and the materials. There were two test fields in
Croatia on the same road, with the same materials, but with application three years apart and at different lines. Weight-
adjusted Annually Averaged Daily Traffic (AADT) was calculated according to Austrian standard ONR 22440-1
(Austrian Standards Institute, 2010), where one heavy vehicle (comprising all buses and all vehicles over 3,500 kg
gross vehicle weight rating) counts as eight light vehicles.
Table 2. Characteristics of the test stretches and applied materials.
Test field
Country, road,
speed limit
Stretch
length [km]
AADT
Base layer,
applied mass
Glass beads,
drop-on mass
All
Heavy
Weight-adjusted
1: all lines
Switzerland,
T5, 120 km/h
0.4
34,965
1,573
45,976
Premium cold plastic,
2.2 kg/m²
100% premium,
0.45 kg/m²
2a: middle lines,
2b: edge lines
Croatia,
D10, 100 km/h
23.2
12,204
986
19,106
Cold plastic,
2.2 kg/m²
30% premium,
0.35 kg/m²
3: middle lines
Poland,
A4, 140 km/h
12.9
18,041
3,703
43,962
Cold plastic,
2.5 kg/m²
Standard,
0.35 kg/m²

Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT. 7
4.3. Results of field evaluations
Results for the test stretch in Switzerland are being also presented elsewhere (Burghardt, 2018). Herein, additional
analysis and comparison with other systems is furnished. Retroreflectivity measured on all of the test fields are
provided in Table 3. The results demonstrate clearly that with the use of 100% premium glass beads, notably higher
initial RL was achieved (average 949 mcd/m²/lx) than with 30% premium glass beads (average 482 mcd/m²/lx), which
was only slightly higher than obtained for standard system (average 438 mcd/m²/lx, which was outstanding result,
indeed). The differences after three years were proportionate to the initially recorded, based on both the time scale
and the vehicular traffic scale, which accounted for much lower AADT in Croatia. Whereas we repeat here that
different traffic loads, climatic conditions, and used cold plastic could have played significant role, the results are
consistent across countries and the trend is maintained.
Table 3. Retroreflectivity at the test stretches.
Test
field
Line
Retroreflectivity (RL) [mcd/m²/lx](a)
Initial
1 year
2 years
3 years
4 years
5 years
6 years
1
Edge right
1010 (153)
567 (127)
547 (163)
446 (93)
341 (52)
–
–
1
Middle
853 (96)
374 (67)
374 (92)
335 (46)
232 (20)
–
–
1
Edge left
984 (151)
688 (62)
675 (120)
518 (94)
372 (36)
–
–
2a
Middle eastbound
600 (66)
n/a(b)
n/a(b)
289 (42)
273 (51)
208 (44)
215 (33)
2a
Middle westbound
545 (61)
n/a(b)
n/a(b)
180 (51)
174 (47)
177 (59)
152 (25)
2b
Edge right eastbound
524 (168)
410 (91)
357 (97)
388 (99)
–(c)
–
–
2b
Edge left eastbound
402 (55)
385 (57)
317 (54)
325 (56)
–(c)
–
–
2b
Edge right westbound
418 (137)
350 (93)
257 (70)
292 (68)
–(c)
–
–
2b
Edge left westbound
404 (95)
366 (85)
257 (63)
295 (69)
–(c)
–
–
3
Middle eastbound
484 (98)
n/a(b)
158 (53)
–
–
–
–
3
Middle westbound
391 (70)
n/a(b)
149 (44)
–
–
–
–
(a)Average of dynamic readings per entire stretch, standard deviations provided in parentheses. (b)Measurements were not done. (c)Testing
continues.
Predictions of durability (length of service before RL<150 mcd/m²/lx), based on passes per carriageway of the
weight-adjusted AADT, are given in Table 4. To estimate the durability, exponential line fit for the entire test period
was used (Abboud and Bowman, 2002); correlation with R²>0.8 based on least squares analysis, except for one case,
was calculated. Because measurements were done only once per year, multiple piecewise prediction model could not
be applied to distinguish between summer (road traffic only) and winter (road traffic and snow ploughs) effects
(Hummer et al., 2011). Much larger RL loss during the first year of usage indicated that the analysis should be separated
into at least two fragments, but also might indicate the ‘diminishing returns’ observed by Abu-Lebdeh et al. (2012).
Quite high correlation between the actual and measured RL values, evidenced by high R², may vary for different road
marking materials, as was recently reported by Babić et al. (2019).
The loss of RL was higher for middle lines (lane division within a carriageway) than for edge lines, which is
reasonable and expected because of the higher number of vehicles encroaching on the markings (Craig et al., 2007).
Performance at the edge lines was following the same durability patterns. Unfortunately, we cannot at present obtain
relevant data from Poland for the standard system at edge lines. The poor results obtained with the standard glass
beads in Poland could also be the result of somewhat worse quality of cold plastic, inappropriate coating of glass
beads, inadequate quality of application (particularly bead embedment – but, inappropriate glass bead embedment
would lead to low initial performance, which was not the case), and numerous environmental factors (especially winter
maintenance). At Croatian test field, there was a notable difference of 55 mcd/m²/lx (9%) between middle lines at the
two carriageways; the difference continued for the subsequent years and during the sixth year increased to 63
mcd/m²/lx (29%). Similar difference was measured for the edge lines in the eastbound direction. We are unable to

8 Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT.
pinpoint the reasons for such a difference, but along with occasional lack of RL loss or even its increases after usage
(Cf. Table 3 for second and third year at test field 2b) these inconsistencies can serve as an example of the difficulties
and uncertainties in assessment of horizontal road markings.
Table 4. Predicted durability of the tested road marking systems.
Test
field
Glass beads
Weight-
adjusted
AADT
Line location
Initial RL
[mcd/m²/lx]
Estimated durability [millions of weight-adjusted vehicle
passes per carriageway], based on exponential line fit
RL<300
mcd/m²/lx
RL<200
mcd/m²/lx
RL<150
mcd/m²/lx
Exponential
line fit (R²)
1
100%
premium
45,976
Edge right
1010
33
44
51
0.90
Middle
853
25
35
42
0.81
Edge left
984
35
46
55
0.94
2a
30% premium
19,106
Middle eastbound
600
14
21
27
0.98
Middle westbound
545
10
17
22
0.89
2b
30% premium
19,106
Edge right eastbound
524
18
30
39
0.61
Edge right westbound
418
11
26
37
0.89
Edge left eastbound
402
18
30
39
0.98
Edge left westbound
404
7
17
23
1.00
3
Standard
43,962
Middle eastbound
484
7
12
16
–
Middle westbound
391
4
11
16
–
5. Financial scenarios
One of the key parameters in selection of road marking systems by road administrators and by contractors is
financial assessment. In doing such a valuation for this paper, it was necessary to make numerous assumptions and
educated guesses, because the actual costing always remains a business secret of the manufacturers and applicators;
therefore, the expense scenarios provided herein are to serve only as a guideline. In Table 5, financial scenarios for
two- and three-year durability of the standard system is compared with multi-year durability of a system reflectorised
with 100% premium glass beads. Assumed were increased costs not only for materials, but also for labour. While the
scenarios in Table 5 contain a time scale, it can be converted to the AADT scale to better convey the message; however,
in such a case the effects of winter maintenance should be taken into account; such analysis is beyond the scope of
this work.
Table 5. Financial scenarios based on durability of road marking systems.
Expense allocation
Standard system
Premium system
Labour, fixed costs, profit
15%
20%
Cold plastic (at 2.20 kg/m²)
80%
85%
Glass beads (at 0.45 kg/m²)
5%
20%
Initial expense
100%
125%
Expected durability [years]
2
3
2
3
4
5
6
Relative total cost per 10 years,
assuming 2-year standard durability
100%
–
125%
83%
63%
50%
42%
Relative total cost per 10 years,
assuming 3-year standard durability
–
100%
188%
125%
94%
75%
63%
The results presented in Table 5 clearly indicate that the higher initial expense associated with the purchase of
premium road marking materials may lead to substantial long-term financial savings. It is obvious that the worst choice

Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT. 9
are road marking systems with low durability. The advantage for contractors who won performance-based tenders
should be obvious, but it can be extended also to road administrators and the entire society. Thus, highly durable road
markings can be seen as the relatively inexpensive solution sought to improve road safety in less developed countries.
Inflation and price volatility were not included in the comparison, even though all those skilled in business analyses
would admit that they could play a significant role. The financial scenarios also exclude any indirect effects like those
associated with road safety, even though increased RL was correlated with lower number of crashes (Carlson et al.,
2013). Furthermore, social benefits of increased mobility of senior citizens and thus increased equal access to
infrastructure, caused by increased RL and thus comfort of driving, are not being calculated. Benefits for improved
MV capability are disregarded, too.
6. Discussion and Conclusions
The differences between performance of road marking systems reflectorised with premium, 30% premium, and
standard glass beads were clear in terms of RL and systems durability. It has to be noted that RL achieved with 100%
premium glass beads was significantly exceeding the existing norms, which never call for more than 300 mcd/m²/lx
initially. The presented results indicate that the norms might need to be adjusted to match the newly available
technologies, in agreement with Jaffe and co-workers (2002) who envisaged constant modification of the standards
with technological change to maximise the benefits and to promote further development.
Cradle-to-grave Life Cycle Assessments (LCA) done on thin-layer road markings demonstrated that the main
parameter for environmental friendliness was not the choice of paints, but the durability of the entire road marking
systems (Burghardt et al., 2016a; Cruz et al., 2016). We are not aware of any published LCA directly related to thick-
layer road marking systems but believe that, like in case of thin-layer applications, durability would also be the
controlling environmental parameter. Given enormous expense associated with accidents, the main financial impactor
in analysis of horizontal road markings effectiveness is the number and severity of accidents, but the second impactor
was recently calculated to be also system durability (Pike and Bommanayakanahalli, 2018).
Since market penetration of road markings with very high RL is low, lacking are analyses related to its influence
on road safety, on fatigue and comfort of drivers, and other associated effects. Based on survey results showing that
drivers notice high RL (Burghardt et al., 2017; Pashkevich et al., 2017) and the reported lesser drivers’ stress on well-
marked roads (Horberry et al., 2006; Diamandouros and Gatscha, 2016), meaningful positive effects can be
anticipated. An eye tracking experiment done in the field has shown that horizontal road markings are perceived by
drivers at the level of gazes, with minimal focus (Pashkevich et al., 2018); their observation depends on the distance
from intersection, which agrees with previous simulator reports related to field of view needed for proper steering of
a vehicle (Land and Lee, 1995). Improved quality and RL of markings can be expected to make vehicle positioning
easier.
It was shown that all of the current issues with the use of MV to recognise and properly classify horizontal road
markings are the same deficiencies that impair their human perception. Presented herein road marking systems are
meeting and exceeding the expectations of the human drivers and MV. Therefore, their maintenance at highest
standards would serve not only human drivers, but also AV, which was recently recognised by European Commission
(2018) in the proposed Third Mobility Package: “Member States shall ensure that road markings and road signs are
properly designed and maintained in such a way that they can be easily and reliably recognised by both human drivers
and vehicles equipped with driver assistance systems or higher levels of automation.” As an additional issue, one
ought to mention the recognition, by both human drivers and by machines, of pedestrian crossings, which belong to
the locations where accidents involving vulnerable road users are occurring very frequently. The same technology for
premium glass beads was recently demonstrated to increase RL and durability, based on a field experiment in
Switzerland (Burghardt at al., 2019a).
In the foreseeable future, horizontal road markings are going to guide not only humans driving their vehicles, but
also computer-controlled automobiles. Human drivers must be able to take over steering of self-driving vehicles until
full automation becomes standard on all vehicles and all roads; therefore, maintenance of horizontal signalisation
infrastructure at appropriately high level is an absolute necessity and the recently developed technology for premium
glass beads that can furnish RL meaningfully beyond the current norms is meeting such requirement. Since road safety
is significantly influenced by horizontal road markings and high-end premium solutions provide excellent results in

10 Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT.
terms of RL, durability, overall price, and environmental friendliness, their use as the first choice for horizontal road
markings on primary roads would be reasonable. Because horizontal road markings are inexpensive and effective
safety features, they are one of the easy to implement solution sought for countries with disproportionally high accident
rates. Road marking tapes, which furnish excellent durability and high RL can also be used, but the expense of their
purchase and application is very high; therefore, they do not meet the requirement of easily available inexpensive
solution.
References
Abboud, N., Bowman, B. L., 2002. Cost-and longevity-based scheduling of paint and thermoplastic striping. Transportation Research Record:
Journal of the Transportation Research Board 1794, 55–62; https://doi.org/10.3141/1794-07.
Abu-Lebdeh, G., Al-Omari, B. H., Ahmed, K., Long, D., 2012. Modeling the impact of winter maintenance on pavement marking retroreflectivity.
Procedia Engineering 50, 942–956; https://doi.org/10.1016/j.proeng.2012.10.102.
Austrian Standards Institute, 2010. ONR 22440-1: Road markings ― Durability ― Part 1: General [in German: ONR 22440-1: Bodenmarkierungen.
Funktionsdauer ― Teil 1: Allgemeines]. Austrian Standards Institute: Vienna, Austria.
Babić, D., Burghardt, T. E., Babić, D., 2015. Application and Characteristics of Waterborne Road Marking Paint. International Journal for Traffic
and Transport Engineering 5, 150–169; https://doi.org/10.7708/ijtte.2015.5(2).06.
Babić, D., Ščukanec, A., Babić, D., Fiolić, M., 2019. Model for Predicting Road Markings Service Life. Baltic Journal of Road and Bridge
Engineering 14.3, 341–359; https://doi.org/10.7250/bjrbe.2019-14.447.
Bektas, B. A., Gkritza, K. Smadi, O., 2016. Pavement Marking Retroreflectivity and Crash Frequency: Segmentation, Line Type, and Imputation
Effects. Journal of Transportation Engineering 04016030; https://doi.org/10.1061/(ASCE)TE.1943-5436.0000863.
Burghardt, T. E., 2018. High durability – high retroreflectivity solution for a structured road marking system. In Proceedings of International
Conference on Traffic and Transport Engineering; Belgrade, Serbia, 27-28 September 2018: pp. 1096–1102.
Burghardt, T. E., Babić, D., Babić, D., 2016. Application of Waterborne Road Marking Paint in Croatia: Two Years of Road Exposure. In
Proceedings of International Conference of Transport and Traffic Engineering; Belgrade, Serbia, 24-25 November 2016: pp. 1092–1096.
Burghardt, T. E., Pashkevich, A., Żakowska, L., 2016a. Influence of Volatile Organic Compounds Emissions from Road Marking Paints on Ground-
level Ozone Formation: Case Study of Kraków, Poland. Transportation Research Procedia 14, 714–723;
https://doi.org/10.1016/j.trpro.2016.05.338.
Burghardt, T. E., Pashkevich, A., Piegza, M., 2017. Drivers’ Perception of Horizontal Road Marking with High Retroreflectivity [in Polish:
Percepcja przez kierowców poziomego oznakowania dróg o wysokiej odblaskowości]. Transport Miejski i Regionalny 8, 5–10.
Burghardt, T. E., Pashkevich, A., Fiolić, M., Żakowska, L., 2019. Horizontal Road Markings with High Retroreflectivity: Durability, Environmental,
and Financial Considerations. Advances in Transportation Studies: An International Journal 47, 49–60.
Burghardt, T. E., Pashkevich, A., Mosböck, H., 2019a. Yellow pedestrian crossings: from innovative technology for glass beads to a new
retroreflectivity regulation. Case Studies on Transport Policy 7.4, 862–870; https://doi.org/10.1016/j.cstp.2019.07.007.
Burns, D., Hodson, N., Haunschild, D., May, D., 2006. Pavement marking photometric performance and visibility under dry, wet, and rainy
conditions: Pilot field study. Transportation Research Record: Journal of the Transportation Research Board 1973, 113–119;
https://doi.org/10.3141/1973-16.
Calvi, A., 2015. A study on driving performance along horizontal curves of rural roads. Journal of Transport Safety and Security 7.3, 243–267;
https://doi.org/10.1080/19439962.2014.952468.
Carlson, P., 2017. Pavement Markings for Machine Vision Systems. Presented at ITS World Congress 2017; Montréal, Canada, 29 October-2
November 2017.
Carlson, P. J., Park, E., Kang, D. H., 2013. Investigation of longitudinal pavement marking retroreflectivity and safety. Transportation Research
Record: Journal of the Transportation Research Board 2337, 59–66; https://doi.org/10.3141/2337-08.
Carlson, P. J., Avelar, R. E., Park, E. S., Kang, D. H., 2015. Nighttime Safety and Pavement Marking Retroreflectivity on Two-Lane Highways:
Revisited with North Carolina Data. In Proceedings of Transportation Research Board 94th Annual Meeting; Washington, DC, United States,
11-15 January 2015: paper 15-5753.
Carlson, P. J., Poorsartep, M., 2017. Enhancing the Roadway Physical Infrastructure for Advanced Vehicle Technologies: A Case Study in
Pavement Markings for Machine Vision and a Road Map Toward a Better Understanding. In Proceedings of Transportation Research Board
96th Annual Meeting; Washington, DC, United States, 8-12 January 2017: paper 17-06250.
Craig, W. N., Sitzabee, W. E., Rasdorf, W. J., Hummer, J. E., 2007. Statistical validation of the effect of lateral line location on pavement marking
retroreflectivity degradation. Public Works Management and Policy 12.2, 431–450; https://doi.org/10.1177/1087724X07308773.
Cruz, M., Klein, A., Steiner, V., 2016. Sustainability assessment of road marking systems, Transportation Research Procedia 14, 869–875;
https://doi.org/10.1016/j.trpro.2016.05.035.
Davies, C., 2017. Effects of Pavement Marking Characteristics on Machine Vision Technology. In Proceedings of Transportation Research Board
96th Annual Meeting; Washington DC, United States, 8-12 January 2017: paper 17-03724.
Diamandouros, K., Gatscha, M., 2016. Rainvision: The Impact of Road Markings on Driver Behaviour-Wet Night Visibility. Transportation
Research Procedia 14, 4344–4353; https://doi.org/10.1016/j.trpro.2016.05.356.

Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT. 11
Dingus, T. A., Guo, F., Lee, S., Antin, J. F., Perez, M., Buchanan-King, M., Hankey, J., 2016. Driver crash risk factors and prevalence evaluation
using naturalistic driving data. Proceedings of the National Academy of Sciences 113.10, 2636–2641; https://doi.org/10.1073/pnas.1513271113.
European Commission, 2018. Proposal for a Directive of the European Parliament and of The Council amending Directive 2008/96/EC on road
infrastructure safety management. COM(2018) 274 final, 2018/0129 (COD). European Commission: Brussels, Belgium. Available at
https://ec.europa.eu/transport/sites/transport/files/3rd-mobility-pack/com20180274-proposal_en.pdf (accessed 22 May 2018).
European Committee for Standardization, 2012. European Standard EN 1423:2012. Road marking materials. Drop on materials. Glass beads,
antiskid aggregates and mixtures of the two. European Committee for Standardization: Brussels, Belgium.
European Road Federation, 2015. Marking a road toward a safer future. An ERF Position Paper on how road markings can make our road safer.
European Road Federation: Brussels, Belgium. Available at https://erf.be/publications/marking-the-way-towards-a-safer-future (accessed 15
May 2018).
Gibbons, R. B., Williams, B., Cottrell, B., 2012. The Refinement of Drivers’ Visibility Needs During Wet Night Conditions. Transportation
Research Record: Journal of the Transportation Research Board 2272, 113–120; https://doi.org/10.3141/2272-13.
Hadi, M., Sinha, P., 2011. Effect of Pavement Marking Retroreflectivity on the Performance of Vision-Based Lane Departure Warning Systems.
Journal of Intelligent Transportation Systems 15.1, 42–51; https://doi.org/10.1080/15472450.2011.544587.
Hadi, M., Sinha, P., Easterling, J., 2007. Effect of Environmental Conditions on Performance of Image Recognition-Based Lane Departure Warning
System. Transportation Research Record: Journal of the Transportation Research Board 2000, 114–120; https://doi.org/10.3141/2000-14.
Hagita, K., Mori, K., 2011. Analysis of the influence of sun glare on traffic accidents in Japan. Journal of the Eastern Asia Society for Transportation
Studies 9, 1775–1785; https://doi.org/10.11175/easts.9.1775.
Hauer, E., 2019. On the relationship between road safety research and the practice of road design and operation. Accident Analysis and Prevention
128, 114–131; https://doi.org/10.1016/j.aap.2019.03.016.
Horberry, T., Anderson, J., Regan, M. A., 2006. The possible safety benefits of enhanced road markings: a driving simulator evaluation.
Transportation Research Part F: Traffic Psychology and Behaviour 9.1, 77–87; https://doi.org/10.1016/j.trf.2005.09.002.
Hummer, J. E., Rasdorf, W., Zhang, G., 2011. Linear mixed-effects models for paint pavement-marking retroreflectivity data. Journal of
Transportation Engineering 137.10, 705–716; https://doi.org/10.1061/(ASCE)TE.1943-5436.0000283.
Jaffe, A. B., Newell, R. G., Stavins, R. N., 2002. Environmental policy and technological change. Environmental and Resource Economics 22.1,
41–70; https://doi.org/10.1023/A:1015519401088.
Land, M. F., Lee, D. N., 1995. Which parts of the road guide steering? Nature 377, 339–340; https://doi.org/10.1038/377339a0.
Matowicki, M, Přibyl, O, Přibyl, P., 2016. Analysis of possibility to utilize road marking for the needs of autonomous vehicles. Smart Cities
Symposium Prague; Prague, Czech Republic, 26-27 May 2016; https://doi.org/10.1109/SCSP.2016.7501026.
Migletz, J., Graham, J., Harwood, D., Bauer, K., 2001. Service life of durable pavement markings. Transportation Research Record: Journal of the
Transportation Research Board 1749, 13–21; https://doi.org/10.3141/1749-03.
Miller, T. R., 1992. Benefit–cost analysis of lane marking. Transportation Research Record: Journal of the Transportation Research Board 1334,
38–45.
Mitra, S., 2014. Sun glare and road safety: An empirical investigation of intersection crashes. Safety Science 70, 246–254;
https://doi.org/10.1016/j.ssci.2014.06.005.
Narote, S. P., Bhujbal, P. N., Narote, A. S., Dhane, D. M., 2018. A review of recent advances in lane detection and departure warning system.
Pattern Recognition 73, 216–234; https://doi.org/10.1016/j.patcog.2017.08.014.
Park, E. S., Carlson, P. J., Porter, R. J., Andersen, C. K., 2012. Safety effects of wider edge lines on rural, two-lane highways. Accident Analysis
and Prevention 48, 317–325; https://doi.org/10.1016/j.aap.2012.01.028.
Parker, N. A., Meja, M. S. J., 2003. Evaluation of performance of permanent pavement markings. Transportation Research Record: Journal of the
Transportation Research Board 1824, 123–132; https://doi.org/10.3141/1824-14
Pashkevich, A., Burghardt, T. E., Żakowska, L., Nowak, M., Koterbicki, M., Piegza, M., 2017. Highly Retroreflective Horizontal Road Markings:
Drivers’ Perception. In Proceedings of International Conference on Traffic Development, Logistics and Sustainable Transport “New Solutions
and Innovations in Logistics and Transportation”; Opatija, Croatia, 1-2 June 2017: pp. 277–287.
Pashkevich, A., Burghardt, T. E., Shubenkova, K., Makarova, I., 2018. Analysis of Drivers’ Eye Movements to Observe of Horizontal Road
Markings ahead of Intersections. In: Varhelyi A., Žuraulis V., Prentkovskis O. (eds) Vision Zero for Sustainable Road Safety in Baltic Sea
Region, Proceedings of the International Conference “Vision Zero for Sustainable Road Safety in Baltic Sea Region”. Vilnius, Lithuania, 5-6
December 2018: pp. 1–10; https://doi.org/10.1007/978-3-030-22375-5_1.
Pike, A. M., Bommanayakanahalli, B., 2018. Development of a Pavement Marking Life Cycle Cost Tool. In Proceedings of Transportation Research
Board 97th Annual Meeting; Washington, DC, United States, 7-11 January 2018: paper 18-1533.
Plainis, S., Murray, I., Pallikaris, I., 2006. Road traffic casualties: understanding the night-time death toll. Injury Prevention 12.2, 125–138;
https://doi.org/10.1136/ip.2005.011056.
Pocock, B. W., Rhodes, C. C., 1952. Principles of glass-bead reflectorization. Highway Research Board Bulletin 57, 32-48.
Räsänen, M., Horberry, T., 2006. Harmonisation of road signs and markings on the Trans-European Road Network to improve road safety in the
EU. Presented at 2006 European Transport Conference; Strasbourg, France, 18-20 September 2006.
Shinar, D., Dewar, R. E., Summala, H., Zakowska, L., 2003. Traffic sign symbol comprehension: a cross-cultural study. Ergonomics 46.15, 1549–
1565; https://doi.org/10.1080/0014013032000121615.

12 Burghardt et al./ Transportation Research Procedia, 48, 3622–3633; https://doi.org/10.1016/j.trpro.2020.08.089. PRE-PRINT.
Sitzabee, W. E., Hummer, J. E., Rasdorf, W., 2009. Pavement marking degradation modeling and analysis. Journal of Infrastructure Systems 15.3,
190–199; https://doi.org/10.1061/(ASCE)1076-0342(2009)15:3(190).
Snowden, R. J., Stimpson, N., Ruddle, R. A., 1998. Speed perception fogs up as visibility drops. Nature 392.6675, 450;
https://doi.org/10.1038/33049.
State of California, 2017. Implementation of six-inch wide traffic lines and discontinuing use of non-reflective raised pavement markers.
Department of Transportation memorandum, 19 July 2017. State of California, Department of Transportation: Sacramento, CA, United States.
Available at http://www.dot.ca.gov/trafficops/policy/memo_6-in-wide-traffic-lines_051917.pdf (accessed 18 June 2018).
Steyvers, F. J., De Waard, D., 2000. Road-edge delineation in rural areas: effects on driving behaviour. Ergonomics 43.2, 223–238;
https://doi.org/10.1080/001401300184576.
Stroila, M. N., Chen, X., Moghaddam, M. K., Lu, V., Kohlmeyer, B. D., 2014. Determining travel path features based on retroreflectivity. United
States Patent No. 8,699,755. United States Patent and Trademark Office: Washington, DC, United States.
Sun, D., El-Basyouny, K., Kwon, T. J., 2018. Sun Glare: Network Characterization and Safety Effects. Transportation Research Record: Journal
of the Transportation Research Board 036119811878414; https://doi.org/10.1177/0361198118784143.
Theeuwes, J., Alferdinck, J. W., Perel, M., 2002. Relation between glare and driving performance. Human Factors 44.1, 95–107;
https://doi.org/10.1518/0018720024494775.
Toroyan, T., Peden, M., Iaych, K., Krug, E., 2013. More action needed to protect vulnerable road users. The Lancet 381.9871, 977–979;
https://doi.org/10.1016/S0140-6736(13)60606-6.
Wang, C., Wang, Z., Tsai, Y., 2016. Piecewise Multiple Linear Models for Pavement Marking Retroreflectivity Prediction Under Effect of Winter
Weather Events. In Proceedings of Transportation Research Board 95th Annual Meeting; Washington, DC, United States, 10-14 January 2016:
paper 16-2425.
World Health Organization, 2016. Global Status Report on Road Safety 2015. World Health Organization: Geneva, Switzerland.
Zhang, Y., Wu, D., 2006. Methodologies to predict service lives of pavement marking materials. Journal of the Transportation Research Forum
45.3, 5–18; https://doi.org/10.5399/osu/jtrf.45.3.601.
Zwahlen, H., Schnell, T., 1999. Visibility of road markings as a function of age, retroreflectivity under low-beam and high-beam illumination at
night. Transportation Research Record: Journal of the Transportation Research Board 1692, 152–163; https://doi.org/10.3141/1692-16.
Żakowska L., 1997. Dynamic Road View Research for Road Safety and Aesthetics Evaluation. Journal for Geometry and Graphics 1, 51–57.