Experimental road signs in a simulated environment research programme – experiment procedures and assumptions

The application of experimental signs on public roads may cause some risks due to driver misunderstanding or misinterpretation, especially when seen for the first time. To minimize this risk, driver reaction and sign understanding can be tested in a safe and relatively cheap simulation-based environment. The consortium running the project “Experimental road marking and its effect on road user behaviour” suggested a similar methodology. The project included a simulation-based research program. The consortium proposed a set of experimental signs which have a strong potential for improving road safety and traffic conditions. The paper outlines the simulation-based research programme involving experimental signs and a proposed experiment procedure. Simulation scenarios and procedures were selected to achieve the required message and goals of implementation. In addition, an analysis is proposed of the factors and indicators of each sign and its effects.


Introduction
Experimental signs are generally used to draw the driver's attention to a new problem or to restrict a certain behaviour.Because experimental signs are not commonly used, drivers may be faced with a sign for the first time in their life.Therefore, experimental signs on public roads may cause some risks due to driver misunderstanding or misinterpretation, affect traffic flow and lead to dangerous situations on the road.To minimize this risk and understand its influence, a simulated driving environment is used before the new signs are deployed in a real road environment.The consortium running the "Experimental road marking and its effect on road user behaviour" project, suggested a similar methodology.In the project, simulationbased experiments for the assessment of experimental signs will be introduced.
This approach was inspired by the promising results of other research teams that used simulation environments to assess how road signs influence the driver.Research conducted by Charlton [1], Rosey and Auberlet [2] or Ding et al. [3] shows that a driving simulation environment is suitable for conducting such research.Using the simulator we can evaluate driver behaviour in a safe environment and see whether the objectives underlying the establishment of a sign will be achieved.The research and the authors' earlier work were the basis for proposing such a methodology.The procedures and assumptions of the proposed research are described in the chapters that follow.

Experimental signs chosen for the simulation-based survey
During the first phase of the project members of the Project consortium selected several experimental signs that should be the subject of the survey during the project.These signs included both horizontal signing and markings and vertical signs.Based on literature review and research results of other international research teams, the selected signs were tested in a simulation-based environment.
The following sign types were selected: -horizontal markings which are the optical means of speed reduction at entries to intersections, junctions and ramp lanes -horizontal curve markings with additional warning signs -horizontal markings for speed reduction -markings indicating the need to keep a safe distance -extension of non-standard horizontal curve markings.Some of the above signs are tested based on good practices and implementation models from other counties.The simulation-based experiment is in this case conducted to verify the influence of the sign on Polish drivers.Other signs, however, have different models of implementation in different countries.Especially in the case of vertical signs, a similar meaning of the sign could be shown using a different pictogram design in different countries.In such cases, the consortium agreed to perform survey-based research, to choose the most understandable sign pictogram.

Survey-based research to choose sign pictograms
Selecting the most understandable sign pictogram is one of the key elements when preparing for a simulation experiment.The drivers who will take part in the study will see these signs for the very first time.The meaning of the sign should be clear, and its comprehension should not influence their driving performance.
Based on literature review, the open-ended survey method based on ISO 9186-1 [4] was proposed.The ISO 9186-1 standard is dedicated to testing non-verbal presentation of information in public areas and specifies methods for assessing the comprehensibility of graphical symbols [4].The method is used for evaluating the share of people who can correctly understand the meaning of a symbol or sign.
At least three designs of signs for each of the meanings were used in the survey.The designs were prepared according to ISO 7010 [5], dedicated to the preparation of safety signs in public areas, and in accordance with national law on graphic symbols to be used in road signs in Poland [6].
For each sign two questions are provided: -What action should you take in response to this symbol?-What do you think is the meaning of this symbol?
The respondent should answer both questions with a few words describing what they thought first when they saw the symbol.Figure 1 shows an example from the questionnaire.All the above parameters would be registered and analysed during simulation-based research.For each parameter the analysis of the behaviour will be conducted on the section before the sign, in the area of the sign's impact and on the section after the sign.
For each of the signs a suitable road situation will be presented.The driver will drive in two similar road situations, one with the experimental sign, and one without it or with the standard road sign.Figure 4 shows an example of an experimental road situation.The process of implementing new road signs must take road safety as a priority.The influence of experimental signs on drivers depends on different contexts (e.g.cultural, organizational, legal, etc.) and needs to be carefully examined before the proposed signs are introduced on the roads.Surveys and a good class driving simulator seem to be the best solution for this purpose due to the controlled and repeatable simulation environment as well as precise driving data registered during the drive.
In the "Experimental road marking and its effect on road user behaviour" project a detailed research procedure was designed in order to deal with the presented topic.The method is based on the literature, international standards and good practices of members of the Project Consortium and provides a two-step procedure for choosing the best-fitting road signs.A simulation-based driving experiment, preceded with selecting the best signs evaluated in an online survey, will provide a sufficient amount of data to evaluate the change in driver's behaviour when faced with the new signs.This procedure is easy to implement and cost-efficient and could significantly simplify the process of road sign implementation, taking into account all accompanying factors.

Fig. 1 . 2
Fig. 1.Example of the survey showing a proposed sign design -horizontal road markings indicating the need to keep a safe distance to the vehicle in front.Every response was categorized into one of five possible categories depending on the degree of correctness.The survey is planned for at least 300 respondents.The results should suggest which of the proposed sign designs is easiest to understand.The signs were used in simulation-based experiment.

Fig. 2 .2. 1 Fig. 3 .
Fig. 2. Example of a Simulation Laboratory research procedure used for experiments divided into two test rides.The procedure applied in Simulation Laboratory helps to control the occurrence of simulation sickness and takes into account the necessity for the driver to adapt their skills to the specifics of driving a car in a simulation environment.A similar procedure was implemented successfully in a previous research e.g.EYEVID [7] and GEMS [8,9].

Fig. 4 .4003 2 . 3
Fig. 4. Example of an experimental road situation.Left: road curve with classical signing.Right: road curve with an experimental sign supplementing the classical sign.