SIXTH FRAMEWORK PROGRAMME PRIORITY 1.6. Sustainable ... · SIXTH FRAMEWORK PROGRAMME PRIORITY 1.6. Sustainable Development, Global Change and Ecosystem 1.6.2: Sustainable Surface

SIXTH FRAMEWORK PROGRAMME

PRIORITY 1.6. Sustainable Development, Global Change and Ecosystem

1.6.2: Sustainable Surface Transport

506716

Best practice guide on road signing

Deliverable No. D 5.1

Workpackage No. WP 5 Workpackage Title Implementation priorities and policy recommendations

Activity No. A 5.4 Activity Title Best practice guide on horizontal, vertical and telematic signing

Author Frank Sulzmann (University of Stuttgart, IAT)

Status: (F: final; D: draft; RD: revised draft)

F

File Name: IN-SAFETY_Deliverable_5.1.doc

Project start date and duration 01 February 2005, 36 Months

IN-SAFETY Deliverable No. 5.1 Dissemination Level: PU Contract N. 506716

January 2008 - 2 - University of Stuttgart, IAT

Table of contents Table of contents......................................................................................................................................2 List of figures...........................................................................................................................................3 Abbreviation List .....................................................................................................................................4

EXECUTIVE SUMMARY .................................................................................................... 5

1 Introduction .................................................................................................................. 6

2 Guidelines ..................................................................................................................... 7

2.1 Horizontal Signing ...........................................................................................................7 2.1.1 Fog speed control .......................................................................................................................7 2.1.2 Black spots.................................................................................................................................7

2.2 Vertical Signing................................................................................................................8 2.2.1 Systematic organization of messages...........................................................................................8 2.2.2 Standardize information provided by signs..................................................................................9 2.2.3 Uniform pictograms..................................................................................................................10 2.2.4 Verbal messages.......................................................................................................................11 2.2.5 Traffic typefaces.......................................................................................................................13

2.3 Telematic signing ...........................................................................................................16 2.3.1 Technical specifications of variable message signs (VMS) ........................................................16 2.3.2 In-car traffic signing.................................................................................................................17 2.3.3 Bilingual messages...................................................................................................................18 2.3.4 Signing on road works..............................................................................................................18

2.4 ADAS..............................................................................................................................20 2.4.1 Timing of queue warning messages...........................................................................................20 2.4.2 Alarm sounds ...........................................................................................................................20 2.4.3 Tactile stimulus combined with alarm sounds ...........................................................................21 2.4.4 Speech versus sounds ...............................................................................................................21 2.4.5 Customization ..........................................................................................................................22 2.4.6 Virtual rumble strips versus rumble strips .................................................................................22

2.5 Traffic planning .............................................................................................................23 2.5.1 Use of Micro-Simulation Models ..............................................................................................23 2.5.2 IN-SAFETY Risk Analysis Method ..........................................................................................24

3 Conclusions................................................................................................................. 25

3.1 Horizontal signing..........................................................................................................25

3.2 Vertical signing ..............................................................................................................25

3.3 Telematic signing and ADAS.........................................................................................25

3.4 Traffic planning .............................................................................................................26

References........................................................................................................................... 27



List of figures

Figure 1: Recommended signs for fog speed control ...........................................................................7 Figure 2: Park & ride, shown with local metro carrier logo and explanatory text ... Error! Bookmark

not defined. Figure 3: Park & ride, without local metro carrier logo ........................................................................8 Figure 4: Evaluated variants (oncoming illegal traffic / wrong way driver) .....................................10 Figure 5: Animated warning sign (oncoming illegal traffic / wrong way driver) .............................10 Figure 6: Typeface Tern .................................................................................. Error! Bookmark not defined. Figure 7: Typeface Tern Greek.................................................................................................................13 Figure 8: Tern for VMS 24 px....................................................................................................................13 Figure 9: Correct answers for testfonts in normal display type display type. DIN is averaged

over the series ...................................................................................................................................14 Figure 10: Correct answers for testfonts in VMS display type. DIN is averaged over the series 14 Figure 11: Traffic typefaces......................................................................................................................15 Figure 12: Viewing distances and point of disappearance.................................................................16 Figure 13: VMS basic layout grid, 1U = 22mm ......................................................................................17 Figure 14: Maximum number of elements on VMS, using flashing triangle..................................17 Figure 15: Lane separation provided by lines according to road markings on road surface .....17 Figure 16: Sample layout long-term roadwork......................................................................................19



Abbreviation List

Abbreviation/Term Definition

ADAS Advanced Driver Assistance Systems D Deliverable

DATEX A network based on X.25, DATEX is an acronym for Data Exchange

DLC Distance Lane Crossing DRAM Darmstadt Risk Analysis Method EU European Union GL Guideline GPRS General Packet Radio Service GPS Global Positioning System HMI Human Machine Interface IIID International Institute for Information Design IN-SAFETY INfrastructure and SAFETY INRAM In-Safety Risk Analysis Method IVIS In-Vehicle Information Systems LED light emitting diode LDW Lane Departure Warning NDMPD Numerically Described Multidimensional Probability Distributions P Parking P+R Park and ride px Pixel RDS-TMS Radio Data System - Traffic Message Channel Tern Typeface, acronym stands for: “Trans European Road Network” TLC Time to Lane Crossing TTC time-to collision TIS Transport Information Service TMC Traffic Management Centre USTUTT University of Stuttgart VMS Variable Message Signs

X.25 a standard protocol suite for connection to packet switched wide area networks using leased lines

WiFi Wireless Fidelity WP Work Package



EXECUTIVE SUMMARY This document combines results from IN-SAFETY WP2, WP3, WP4, WP5 and takes into account previous knowledge from other projects to result in a best practice guide. It gives an overview about horizontal, vertical and telematic signing. Every article deals with only one subject without going too much into detail. All topics are explored from different points of view. Often information from different sources are brought together. In some cases pilot studies of IN-SAFETY challenge the existing best practice.

Most chapters provide information about today’s best practice whereas some chapters give an outlook and present ideas having high potential to be best practice in near future.

The best practice guideline is divided in five sections and covers following topics: • Horizontal signing:

o Using high performance retro-reflective technology to reduce accident rates o Inventive ways using horizontal signing

• Vertical signing: o Suggestions and a systematic approach to harmonize symbols, keywords o Introduction of a new traffic typeface available for 25 languages, for all countries

in EU o Standardizing information o Icons o Verbal messages

• Telematic signing o Look and feel of future VMS (variable message signs)

• ADAS:

o How to improve existing ADAS (Advanced Driver Assistance Systems) o In-car traffic signing o Is there potential to substitute the infrastructure element milled rumble strips by an

in-vehicle assistance system? • Traffic planning

o Using micro- and macro simulation o Risk analysis



1 Introduction The IN-SAFETY project aims to enhance the forgiving and self-explanatory nature of roads by creating intuitive and cost-efficient combinations of new technologies and traditional infrastructure best-practice applications. In order to accomplish this, the project incorporated:

• Building consensus on priorities for regulation and standardisation processes and assessing the potential and cost-effectiveness of combined use of such new technologies (advanced driver aid systems, ADAS; in-vehicle information systems, IVIS) and innovative human machine interface (HMI) concepts.

• Developing and testing new simulation models and risk analysis tools, to estimate the safety of road environments.

• Developing training tools and curricula for road and traffic management centre (TMC) operators, focusing on the use of new technologies.

• Optimising vertical and horizontal signing and personalising their information to the specific needs and wants of each user.

• Issuing priority implementation scenarios, guidelines for further research and policy recommendations for cost-efficient road environment development, road safety assessment and inspection, including new technological elements.

This document combines results from IN-SAFETY WP2, WP3, WP4, WP5 and takes into account previous knowledge from other projects to result in a best practice guide. It gives an overview about horizontal, vertical and telematic signing. Every article deals with only one subject without going to much into detail. If you are interested in further information please have a look at the respective sources.

The best practice guideline is divided in main five sections: • Horizontal signing • Vertical signing • Telematic signing • ADAS • Traffic planning

All topics are explored from different points of view. Often information from different sources are brought together. In some cases pilot studies of IN-SAFETY challenge the existing best practice.



2 Guidelines

2.1 Horizontal Signing

2.1.1 Fog speed control Fog speed control can be achieved through information given on boards positioned along the motorway. Speed adjustment in foggy conditions can be done according to white dots painted in appropriate intervals on the side of the motorway. If two dots can be seen at the same time, a recommended speed reduction to 60 km/h is advised. If only one dot can be seen, drivers are advised to reduce the speed to a maximum of 40 km/h. Certain weather conditions like snow may require even lower speeds, but this does not conflict with existing rules.

This requires of course that in districts of frequently occurring fog white dots are painted in appropriate intervals on the side of the motorway. [1]

This is an alternative to The German Road Traffic Regulation (StVO): If visibility is less than 50 metres, the driver must not drive faster than 50 kilometres per hour. For a quite exact estimation of a 50 metre-sight in Germany, delineators besides the road are placed usually at intervals of 50 metres. But drivers have to remember this rule for themselves.

Using traffic signs as shown in Figure 1 has the advantage that speed limits can be adapted to the locality and the specific road situation. These signs are more flexible than the German Road Traffic Regulation (StVO) rule and less expensive than VMS (maintenance and installation).

This kind of fog speed control was first developed in Italy and later on it was adopted by Austria:

Figure 1: Recommended signs for fog speed control

2.1.2 Black spots Black spot situations can be improved dramatically by low cost upgrading of signing and pavement markings. The use of high performance retro-reflective technology has demonstrated in various situations that improved conspicuity of the signs and markings have resulted in lower accident rates. This improved visibility is needed around the clock for all kind of weather conditions, for any age of person driving any type of vehicle.

A summary of some case studies / best practices by various authorities out of various European countries can be found in [14].



2.2 Vertical Signing

2.2.1 Systematic organization of messages A systematization would on the one hand enhance the comprehensibility and perceivability, increasing transparency and robustness. On the other hand systematization would reduce variation of existing signs and messages. Dynamic development or ad-hoc creation of new verbal messages would be simplified without deprecating comprehensibility and perceive-ability.

One would be highly flexible for new developments and new requirements. This would also greatly increase the investment safety in the field of traffic signing.

The results of IN-SAFETY WP 2 based on extensive investigations can be considered as a first step towards systematizing and harmonizing verbal messages across Europe. Due to the promising result of this study following procedure for further systemisation can be recommended: • Harmonizing verbal messages:

quod vide 2.2.2

• Simultaneously systematize the syntax of clusters of co-occurring signs, graphical symbols and verbal messages/verbal message elements, such as: o P (parking) § extension level 1: for cars, buses, trucks... § extension level 2: indication of direction, distance, free parking spaces § extension level 3: indication of additional information: WC, drinks, time

information

Without information about distance and the number of free parking spaces the decision-making of driver is problematic and it can cause waste of time and running additional kilometres which means higher fuel consumption and increased burden to the environment. Here it is obvious that the language independent supplementary information is important. These information has to be put on co-occurring signs (not shown here).

The examples show parking signs with information of extension level 1:

Figure 2: Park & ride, shown with local metro carrier logo and explanatory text

Figure 3: Park & ride, without local metro carrier logo

[1]



2.2.2 Standardize information provided by signs Inter border travelling has become an everyday reality. So it is important that in every country the information is presented in a similar way so that it will be easy for drivers to understand the messages in every country. Even if they are not completely familiar with the language and the road network, a familiar format of the messages will make it easier for them to understand the meaning of it. [8]

Identifying ”key meanings” was carried out after extensive investigations done in IN-SAFETY by selecting a number of verbal messages, which could be considered as candidates for ”Europeanisms” (see 2.2.4) – i.e. verbal messages, which have the potential to be harmonised EU-wide.

They can largely be subdivided into: • units and measures • (short) verbal messages or verbal message elements

Some of the latter can either replace a graphic symbol or verbalize a graphic symbol. In this connection it must be emphasized that Europe-wide harmonization does not necessarily conflict with “multilingualism”, since the individual languages will still be used in most verbal messages – especially those of a more complex nature, whether in navigation systems or in other types of car-driver communication systems.

The results of IN-SAFETY [1] are a base for further discussions. It is recommended to follow this systematic approach to harmonize as many verbal messages and verbal message elements (as well as traffic signs and graphic symbols) as possible.

A catalogue of fundamental verbal message elements considered for Europe-wide harmonization can be found in [1]. This catalogue provides a basis for further discussions.

However, any harmonisation of verbal messages can only be developed as far as it is certain that nearly everybody is able to understand the message. It can be assumed that comprehensibility is best if words are presented in native language. If possible information should be provided both in native language and using fundamental verbal message elements.

[1]



2.2.3 Uniform pictograms A variety of icons is used for the same message across Europe. To facilitate journeys in international corridors it is recommended to uniform pictograms. [7]

Figure 4: Evaluated variants (oncoming illegal traffic / wrong way driver)

In a study of IN-SAFETY over 560 pictograms were tested and elaborated. Basing on these results several pictograms were redesigned. So there is now a set of validated and standardized pictograms available. This set is a starting point for further discussion and analysis. It can be found in [1].

In near future freely programmable VMS will give more options in displaying icons. For example animations and flashing elements can be used. To alert drivers to a rapidly approaching danger it is suggested to superimpose a flashing triangle or a flashing “X” onto the static symbol/pictogram. Test results have proven that superimposing animated pictograms with a flashing graphic element impairs comprehension. Therefore such combinations are recommended only for static symbols to prevent possible distraction of the driver.

Animation of icons improves comprehension only in certain cases. For example to express an action like “oncoming wrong-way-driver”:

Figure 5: Animated warning sign (oncoming illegal traffic / wrong way driver)

Considering the many conditions governing the quick discriminability and correct comprehension of symbols/pictograms, test results indicate correlations between the degrees, to which a pictogram relates to already learnt information. Traffic signs usually have an advantage over newly introduced pictograms. Obviously clear and simple images can be comprehended quicker than detailed ones. These two factors were considered in the elaboration of a concept aiming at weighing pictograms for their use on VMS to safeguard comprehensible messages which do not overburden motorists. [1]

Three classes of symbols/pictograms were defined in [1]: • Class 1

o All Vienna Convention signs required for messages on VMS, unless tested for comprehension with scores below 88

o Pictograms not regulated by the Vienna Convention, that have yielded comprehension scores above 88

o Europeanisms, see [1]



• Class 2 o Vienna Convention signs tested for comprehension with scores between 77 and 88 o Pictograms not regulated by the Vienna Convention, that either have been accepted

after convincing scores when tested for judged comprehensibility or yielded comprehension scores between 77 and 87.

• Class 3 o Vienna Convention signs tested for comprehension with scores between 66 and 77 o Pictograms not regulated by the Vienna Convention, that have yielded comprehension

scores between 66 and 77. Some of these pictograms are considered on the provision that they get regulated, subsequently learnt in driving schools and advertised widely to induce a learning process among drivers.

2.2.4 Verbal messages In principle also all non-linguistic representations may have to be rendered in verbal form in some situation or for some special user groups. Verbal messages may appear in written or spoken form. Non-linguistic representations here refer not only to graphic signs and symbols, but also to other non-linguistic representations, such as haptic information (e.g. vibration of the steering wheel in case of speed limit violation).

In general fundamental verbal messages should be written in upper case letters whereas measures, units and quantities are recommended to be written in lower case letters. The well known “i” for information should be written in lower case, too.

Verbal messages should not exceed 5-6 letters whenever possible (exceptions: EXCEPT, CONTROL, etc.). If applicable be based only on metric measures, units and quantities in Europe.

These fundamental (and short) verbal messages should be recognizable as “icons” rather than to be read. Therefore, a European-wide harmonization would reduce the learning necessity and increase the iconic character of verbal messages.

Needless to say that for the spoken form a different – however coordinated – systematization has to be developed, which will heavily rely on the phonological system of the respective language.

In IN-SAFETY a catalogue of “Europeanisms” (verbal messages, which have the potential to be harmonised EU-wide) was created to serve as a base for further discussions. However, any harmonisation of verbal messages can only be developed as far as it is certain that nearly everybody is able to understand the message. It can be assumed that comprehensibility is best if words are presented in native language. If possible information should be provided both in native language and using fundamental verbal message elements.



Examples:

Restriction: Letter sign Meaning represented FULL e.g. parking area full

NO not permitted, not allowed, denied

OK permitted/allowed

EXCEPT Except for …

t ton (weight limit)

km kilometre (per hour, distance)

STOP “stop” sign

Warning: Letter sign Meaning represented FOG Fog warning

OIL Slippery road due to trail of oil

SMOG polluted air

Information:

Letter sign Meaning represented @ Internet access available

BUS bus

CENTRE / CENTER (city) centre

EXIT exit from highway, building…

GRATIS access/use etc. free of charge

H / HALT halt (bus, tram etc. stop)

i / INFO information (point/available/…)

M / METRO metro, underground

P / PARKING parking (area/building/…)

P+R park-and-ride

POL / POLICE police

RADIO traffic broadcast + frequency

SOS emergency (telephone, vehicle, …)

TAXI e.g. for taxis only

TEL telephone

WC toilet (available…)

Accessory information: Letter sign Meaning represented km kilometre (distance)

min minute

via Reach location A via location B

[1]



2.2.5 Traffic typefaces

Figure 6: Typeface Tern

Figure 7: Typeface Tern Greek

Figure 8: Tern for VMS 24 px In IN-SAFETY the typeface “Tern” was designed according to analysis of the three most influential European traffic typefaces: • DIN (Germany) • RWS (Netherlands) • Transport (United Kingdom)

These typefaces are today’s best practice in the EU and used for VMS and traffic signs.

The typeface “Tern” is available for regular roads sign application (Error! Reference source not found.) and it is able to cater for 20 EU languages, e.g. Greek ( Figure 7). At this stage (January 2008) some European countries are highly interested in the Tern and are evaluating this new typeface for further use: • Austria • Finland • The Netherlands

For VMS special screen fonts in several font sizes are also available (

Figure 8); they are called TernVMS. These bitmap fonts are more than a simple conversion of an outline font to a bitmap font. They are made with big effort and they are specially



optimized for VMS regarding the low resolution of VMS panels. For example the typefaces differ in “Q”, “S”, and “W”.

Tern characters were tested for discriminability in comparison to DIN, RWS and Transport in a special impaired visibility typeface test. Tern scores significantly better than DIN in normal display with respect to the frequency of correct answers. The results for VMS display show no meaningful differences. Under the extended testing conditions of impaired visibility and dual-purpose-display-mode (normal and VMS) an empirically grounded legibility-ranking had been established. A general decline in legibility could be uncovered from Transport to Tern, followed by DIN and the RWS-font for VMS displayed typefaces.

After testing “Tern” typefaces were redesigned according to the test results. So Tern is even better than following results indicate.

Test results:

Figure 9: Correct answers for testfonts in normal display type display type. DIN is averaged over the series



Figure 10: Correct answers for testfonts in VMS display type. DIN is averaged over the series

The results show that Tern scores better than DIN and RWS. Transport scores best, but it needs a lot of space.

It is in question if is sensible to substitute existing traffic signs using the Tern. On one hand discriminability and safety would be enhanced but on the other hand the costs in no way justify the benefits.

It is recommended to use TernVMS on freely programmable VMS using LED’s on a grid of increments of 22 mm, see 2.3.1. So the resolution is optimal and TernVMS can be reproduced correctly. By comparison to other typefaces Tern is compact and very well distinguishable (Figure 11). Therefore TernVMS is most suitable for VMS.

[1]

Figure 11: Traffic typefaces

Caption: 1. Arial narrow 2. MITT2 (Germany) 3. Tern (IN-SAFETY project) 4. Transport (United Kingdom) 5. DRIP font (Netherlands)



2.3 Telematic signing

2.3.1 Technical specifications of variable message signs (VMS) Basically it is the emerging new generation of freely programmable VMS (Variable Message Signs) along with the insight that effective communication often requires the combination of

various information elements, e.g. pictograms with text below or alongside. Fully programmable VMS offer the option to display lane-specific information and to animate the information whenever heightened alertness is on demand.

Today completely freely programmable VMS haven’t been authorized yet (for example in Germany). But there is high potential by using this new kind of VMS. Therefore following statements give an outlook. It is important to know which technical demands have to be met in the near future!

Figure 12: Viewing distances and point of disappearance

The resolution of VMS must allow for the discriminability of the smallest graphical detail of the characters and symbols/pictograms to be displayed. With regard to the distance of 107,06 m from where the VMS information must be perceivable and comprehensible, the smallest graphical detail for eyes of visual acuity 0,73 needs to be about 44 mm in size. Increments of

22 mm are close to many newly manufactured VMS. 22 mm also seems to be a good compromise between other practices which range from 15 mm up to 25, 30 and even 60 mm.

Considering a viewing distance of 14,55 m (point of disappearance) the resulting resolution turns out to be 20 lines/cm = 51 lines/inch. This corresponds to the resolution of pictures on low grade newsprint common at the time after the Second World War.

The width of displays corresponds with the width of 2 lanes, which means approximate 7,50 m. The size of the information depends on driving speed. Due pictograms have to be distinguishable and discriminable they need a minimum height of 64 pixels at a speed of 100 km/h. Using LED’s on a grid of increments of 22 mm means that the height of displays has to be at least 1,40 m. The minimum size for text is 20 pixels, so three lines of text accompanying one pictogram can be used.

At speeds above 100 km/h larger information on larger displays is needed. The amount of information has to be reduced, if width remains unchanged at 7,50 m.



Acceptable deviations: • Reduce pictogram size to 46 pixels (only for well known and easily discriminable traffic

signs) • Reduce letter size to 14 pixels

(only for short, complementing, internationally well understood words, e.g. "100 m") • Letter size enlarged to 31 pixels

(only for short, internationally well understood words, e.g. used for urgent warnings)

Figure 13: VMS basic layout grid, 1U = 22mm

Some information elements such as pictograms are better comprehended when shown in context, e.g. "Road works" + "Speed restriction to 80 km/h".

However the more information is given, the higher the probability that some of it gets neglected…

Figure 14: Maximum number of elements on VMS, using flashing triangle

Figure 15: Lane separation provided by lines according to road markings on road surface

[1]

As you can see all elements are as large as possible to be quickly and largely comprehended. In Figure 14 the worker is displayed in full size and an flashing triangle is used – see 2.2.3. As well the number “80” is enlarged to enhance discriminability.

2.3.2 In-car traffic signing The results suggest road authorities to promote further development of this type of information systems, which seem to be proper to various user groups.

Integration of traffic sign information on an in-car device is a promising approach. It seems to contribute to improve the effects of warning signs and speed limit signs.

For the road authorities it is important to find new approaches to enhance traffic safety. In this context, it also is road authorities’ and operators’ interest that all traffic signs are perceived



and that they affect driver behaviour as meant. In practice, the messages of warning signs are frequently not detected however, and therefore it would be preferable to find out ways to support traffic sign information. [6]

2.3.3 Bilingual messages Bilingual message signs are becoming more and more important throughout the roads of Europe growing together. People are travelling more intensively through foreign countries than ever before. Modern technology makes it possible to introduce bilingual variable message signs.

There are two ways to display bilingual messages: • All languages are displayed simultaneously • Languages are displayed consecutively

Four-line bilingual VMS, comprising two lines of text in each language displayed in turns, have been tested and validated as the most acceptable solution. Messages should consist of less than 6 units of text if the displayed information is supposed to be recalled later.

Regarding visual distraction, it does not make a difference if bilingual messages are displayed consecutively or simultaneously. So finally, it is acceptable to display bilingual messages by turns of two seconds each. It is no more demanding to display variable message signs alternatively than to display them simultaneously. However elderly drivers consider VMS as more demanding.

[1]

2.3.4 Signing on road works Safety is adversely affected at roadworks. It is proven that road work zones present higher accident rates than non-works sections. The safety problem of road work zones deserves special consideration, for the following main reasons: • A work zone signifies a temporary change in the standards of the road facility travelled,

most usually a deterioration, potentially leading to violation of driver expectancies.

• Road work zones occur relatively frequently, since in European and other countries an increasing proportion of highway projects consist of either improving or maintaining existing facilities.

• The accidents that occur at work zones may involve not only road users (drivers, cyclists, pedestrians) but also site personnel.

In ARROWS a best practice handbook was produced for the practical guidance of network managers at all levels. In many cases the recommendations of this handbook differ from the national standards for road work zone safety in European countries. Moreover, the handbook should be consulted in cases where procedures and responsibilities are not adequately defined in national standards.



The following basic principles have been applied: • Only the decisive characteristics of a work zone are shown, presenting a minimum

number of elements for the various traffic routing situations; adaptation to traffic routing situations can be achieved in a schematic manner, by using and exchanging only a small number of basic elements.

• Alternative variants of safety measures are indicated where appropriate.

• Suitable distances have been defined for locating the announcement and other critical points along road work zones.

• Subject to national legal specifications of maximum speed limit values, speed limits along the road work zone are reduced by regular steps (e.g. by 20 km/h), equally-spaced (”speed funnel”).

Layouts provide information about used elements and a logical way to place them giving information on suitable distances such for narrowing area, stabilizing area and transition area. More details and layouts can be found in [16]

Figure 16: Sample layout long-term roadwork



2.4 ADAS

2.4.1 Timing of queue warning messages The timing of an incident / queue warning has a high correlation to the driving situation and environmental aspects. Thus the effect of the information depends on the accuracy of the time interval between warning and event. Here the balance has to be found between presenting the warning early enough to react and late enough not to forget it. [6]

In a pilot study of IN-SAFETY subjects were confronted with two situations with a parked school bus during each drive. One situation without prior warning and one situation with an in-vehicle warning before the school bus became visible. The results of this part of the pilot supported the hypothesis that an in-vehicle warning results in lower speeds when passing the bus compared to a no prior warning situation. Especially the speed development shortly before passing the parked bus was more favourable in terms of traffic safety when the drivers had received the prior warning.

There is consequently a potential to use in-vehicle information to reduce speeds during temporary safety critical events. Possible adaptation effects on such a warning strategy should be examined in a long time study. [5]

2.4.2 Alarm sounds Alarm sounds have to be loud enough to be perceived but otherwise they shouldn’t be harmful of course. Therefore auditory tones should be about 15dB above the masked threshold, but no more than 115dB absolute level. The 15 dB level reflects a compromise of the recommendations found in studies investigating the use of warning sounds in aircraft cockpits. [9]

To create distinguishable sounds, vary two or more of the following parameters: • Spectral content • Pulse duration • Pulse shape • Temporal pattern

The sound should be composed of 10 or more harmonically spaced components, at least 4 of which are prominent and in the range of 100 to 4000Hz. Most of the energy of lower-priority warning signals should be in the first 5 harmonics, whereas higher-priority signals should have relative more energy in harmonics 6 - 10. [10]

The duration of a signal should be between 100 and 150ms. [10]

Problems with alarm sounds identification are very individual – people consider various pitch of tone (or their combination) as a signal of danger. It requires developing of conditioned reflex in the driver. The importance of customization (see chapter 2.4.5) is obvious.



2.4.3 Tactile stimulus combined with alarm sounds Alarms signifying immediate danger should be presented through two sensory modalities to improve their likelihood of being received. Because of differences in drivers' perceptions and differences in driving environments, dual modality alarms are likely to be more effective, especially since the driver is hypovigilant. For example, an auditory tone combined with a haptic stimulus is likely to be more effective over a broad range of drowsy drivers than just a tone.

Subjects participating in driving simulator studies have shown a greater reduction in their level of alertness than subjects involved in road studies. Simulators, for the most part, have a smoother ride and less road vibration than an actual car. This suggests that vibration may have an alerting effect and for this reason, should be considered a viable warning signal for driver drowsiness. It is suggested that tactile displays, such as vibration, is located in the driver's seat, the steering wheel or the seat belt lock. Findings within the AWAKE project show high acceptance and effectiveness rates for warnings incorporating seat belt vibration.

Tests within the AWAKE project have shown vibrations at the seat belt with a frequency of around 125 Hz to be optimal for detection. [15]

2.4.4 Speech versus sounds Use non-speech auditory messages (sounds) only for the purposes of alerting either as a self-contained message, or as a method of alerting the driver to an in-vehicle visual message or to a spoken message that follows.

Other auditory messages, including complex warnings, should be speech. Computer-generated, online speech is recommended for situations that require substantial flexibility in generating spoken messages. Where the choice of messages is relatively limited and economy of time is needed, recorded human speech is preferred. [11]

A tonal signal should be used for attracting attention, and for delivering a preparatory message. It may also be selected to deliver general information such as "attention" or "danger".

A tonal code is also recommended for information requiring a quick response from the driver, but only if the meaning is obvious. The number of tonal signals used in a car should be limited to 3. Tonal codes should be perfectly audible and recognizable without any risk of confusion for the drivers, whatever the characteristics of the driving situation are. [12]

Messages that require an urgent action should be a single word or a short sentence with the fewest number of syllables possible. Drivers should be able to understand the message immediately.

Messages that are not urgent or for which a response may be delayed can be a maximum of 7 units of information in the fewest number of words possible. If the information cannot be presented in a short sentence, the most important information should be presented at the beginning and/or the end of the message. Navigation instructions should be limited to 3 or 4 information units. [13]



2.4.5 Customization Although there are many guidelines and recommendations how to design warning messages, alarm sounds and so on, it is impossible to design alarm messages suitable for all drivers and for every situation. Therefore users should be able to personalize their applications, e.g. adjusting alert levels.

In a pilot study of IN-SAFETY participants rated a headway personalized application more favourably than the baseline application. It was observed that the driver becomes less nervous while using the personalized systems. The warnings were activated fewer times as compared to the warning activation in the system with standard thresholds, thus resulting in the fact that personalized systems increased confidence and promoted higher acceptance by the drivers. When evaluating the standard systems, the number of warnings activated was much more, thus considered to be as less flexible to the driver’s behaviour.

For safety reasons, a minimum threshold for headway and distance from left lane marking should be utilized, thus reducing the warning limits personalization.[4]

2.4.6 Virtual rumble strips versus rumble strips Lateral position is an important issue in relation to rumble strips since shifted average and reduced variance in lateral position has implications for maintenance costs.

Rumble stripes are a physical modification of the road surface. They lead to an acoustical and perceptible effect to warn the driver before the vehicle drifts off the road. “Virtual rumble strips” means a vehicle-guidance-system or a driver-assistance-system, which warns the driver too.

In a pilot study of IN-SAFETY it was shown that there is an high acceptance of both milled and virtual rumble strips. Indeed there is no significant difference between the acceptances of the two rumble strip types. Observations together with the similar driving data observed for milled and virtual rumble strips suggest that there is a potential to substitute the infrastructure element milled rumble strips by an in-vehicle assistance system. No significant shift in lateral position was found.

The fact that virtual (invisible) rumble strips lead to similar driver behaviour than real milled rumble strips speaks in favour of their effect. [5]



2.5 Traffic planning

2.5.1 Use of Micro-Simulation Models Microscopic traffic simulation models create traffic flow as the movements of individual vehicles – just like in reality. It seems obvious that they are used for safety analysis because they show distinct advantages like reproducibility, easy-to-produce large sample sizes and arbitrary, also safety-critical situations without endangering human lives or damaging valuable goods. This is also the reason why simulation techniques have been used more and more in mechanical experiments (e.g. crash tests). The same advantages can be gained with traffic flow simulations.

An important requirement is, however, an optimal calibration of the model. Safety problems normally occur, when rare parameter constellations arise. In statistical terms, this means that such situations only happen when parameters of the “tails” of the distributions are combined. The resulting requirement is a good calibration not only for “average” situations – the distributions from which parameters for driver behaviour etc. are drawn must reflect reality totally. An average driver in an average situation does not cause an accident; the problem lies in the combination of the last few percents in all parameters describing driver, vehicle and surroundings.

Such good calibrations are normally extracted from dedicated experiments like in-vehicle measurements or simulator studies to adequately describe driver behaviour. The traffic situations may be deducted from traffic measurements if these show a suitable resolution to derive distributions. Optimally they consist of individual vehicle data to obtain information on headway distribution etc. The most commonly applied aggregation of data into time intervals provide only indicative results which may be used if more information, like the assumption of the shape of a headway distribution, can complement the information.

Given a well calibrated model, it is possible to derive safety effects in a broader sense than in reality. With the restricted sample sizes of critical situations or accidents, analysis often runs into statistical problems. With micro-simulation this can be overcome: the sample size is normally only restricted by practical limitations, like run-time, but is generally larger than real-life observations.

Furthermore, it is possible to extract safety relevant parameters like time-to collision (TTC), share of short headways, strong decelerations etc., which indicate a safety level on a larger sample size and therefore statistically more significant. Given such surrogate parameters, micro-simulation can better estimate safety effects than other methods relying on a small number of real-world safety-critical data.

More details of this how are explained in [2]



2.5.2 IN-SAFETY Risk Analysis Method Self explanatory roads make user behave correctly without thinking about. To assess, whether a design of a road infrastructure, a regulation or an ADAS-system supports the self explanatory property of the road, it is necessary to integrate human behaviour and acceptance into the evaluation process.

Due to its individualism, it is very difficult to integrate human behaviour into the analysis. It may be done by Numerically Described Multidimensional Probability Distributions (NDMPD). The developed Darmstadt Risk Analysis Method (DRAM) allows the use of such NDMPD (a tool to handle them is provided) and furthermore allows to model wide parts of the road system in an effective and modular manner.

Other parts, taken from ADVISORS project, allow analysing legal and organisational influences. The overall covering methodology is called In-Safety Risk Analysis Method (INRAM).

More information is given in [3]



3 Conclusions

3.1 Horizontal signing In an IN-SAFETY pilot study [5] it was shown, that there is a potential to substitute the infrastructure element milled rumble strips by an in-vehicle assistance system. This will probably be best practice in future and will reduce costs in road maintenance.

The use of high performance retro-reflective technology in combination with fluorescent colours will reduce accident rates. There is still leeway to use markings for special purposes such using markings for fog speed control.

3.2 Vertical signing It is recommended to follow a systematic approach to harmonize pictograms, verbal messages and the systematic organization of message elements. So one would be highly flexible for new developments and new requirements. Furthermore this would also greatly increase the investment safety in the field of traffic signing.

In WP 2 of IN-SAFETY over 560 pictograms were tested and elaborated. Basing on these results several pictograms were redesigned. So there is now a set of evaluated pictograms, unified keywords and a universal typeface (for signing and VMS in 25 languages) available. These results are a base for further discussions and a first step to harmonize symbols and keywords as much as possible.

Generally it is recommended to evaluate pictograms, messages, etc before they are really used in practice

3.3 Telematic signing and ADAS There is a lot of knowledge available to deal with diverse problems. It is important to be up to date and to challenge existing guidelines.

Pilot studies of IN-SAFETY showed that there is high potential in improving existing in-vehicle systems. Systems which are giving prior warnings to critical situations have big positive effects. Customization of warning systems enhances driver’s acceptance and increases confidence.

Freely programmable VMS allow to animate pictograms for possibly further improved comprehension and to optimize them therefore for impaired visibility conditions.



3.4 Traffic planning Microscopic traffic simulation models create traffic flow as the movements of individual vehicles – just like in reality. It is possible to derive safety effects in a broader sense than in reality. With the restricted sample sizes of critical situations or accidents, analysis often runs into statistical problems. With micro-simulation this can be overcome: the sample size is normally only restricted by practical limitations, like run-time, but is generally larger than real-life observations. Furthermore, it is possible to extract safety relevant parameters which indicate a safety level on a larger sample size and therefore statistically more significant. Given such surrogate parameters, micro-simulation can better estimate safety effects than other methods relying on a small number of real-world safety-critical data.

As seen in 2.5.2 there is a risk analysis tool available which allows to work with a huge amount of data and to integrate human behaviour into the analysis. It is not necessary to reduce data to more or less arbitrary characteristic values, which normally over- or underestimate risk systematically. The possibility for quantitative description and evaluation principally allows any desired precision. The precision is only dependent on the availability and the gathering of data. In fact, the traditional methods may be integrated into this methodology.



References [1] P. Simlinger, S. Egger, C. Galinski

“Proposal on unified pictograms, keywords, bilingual verbal messages and typefaces for VMS in the TERN” Deliverable 2.3, C.N. 506716, IN-SAFETY project, January 2008

[2] T. Benz, E. Gaitanidou, I. Spyropoulou, S. Toffolo “Improved Micro and Macro Simulation Models” Deliverable 3.1, C.N. 506716, IN-SAFETY project, February 2007

[3] J. St. Bald, K. Stumpf, T. Wallrabenstein, L.T. Huyen “Road Risk Analysis tools” Deliverable 2.3, C.N. 506716, IN-SAFETY project, January 2008

[4] A. Anund, et al. “Pilot results consolidation, Greek pilot” Deliverable 4.2, C.N. 506716, IN-SAFETY project, January 2008

[5] A. Anund, et al. “Pilot results consolidation, Swedish pilot” Deliverable 4.2, C.N. 506716, IN-SAFETY project, January 2008

[6] K. De Brucker, J. Weinberger, et al “Implementation scenaria and further research priorities”, Guideline 4A39 Deliverable 5.3, C.N. 506716, IN-SAFETY project, January 2008








[14] R. Nuyttens “Blackspot Management: Low cost measures offered by horizontal and vertical signing”, Road Safety on Four Continents Conference, 5-7 October 2005, Warsaw, Poland

[15] E. Bekiaris, et al “Design guidelines for driver drowsiness detection and avoidance”, Guideline 13 and 14 Deliverable 9.1, report IST-2000-28062, Awake project, August 2004

[16] G. Kanellaidis, et al “Road Work Zone Safety Practical Handbook” Deliverable 4, Contract No. RO-96-SC.401, Arrows project, November 1998

Documents

SIXTH FRAMEWORK PROGRAMME PRIORITY 1.6. Sustainable ... · SIXTH FRAMEWORK PROGRAMME PRIORITY 1.6. Sustainable Development, Global Change and Ecosystem 1.6.2: Sustainable Surface