Comparing Driver Behaviour with Measured Speed—An Innovative Approach to Designing Transition Zones for Smart Cities

Abstract

Speed limits are widely used in transition zones between rural and urban areas, where road and environmental conditions change and drivers are expected to reduce their speed. These locations often generate particularly complex driver behaviour in response to applied traffic calming measures (TCMs). Previous studies have mainly focused on the effectiveness of individual TCMs in reducing speed; however, analyses directly comparing drivers’ declared behaviours with actual measured speeds remain limited. The aim of this study was to assess the effectiveness of selected TCMs—chicanes, central island, refuges island, and dynamic speed feedback signs (DSFSs)—across 26 transition zones, taking into account land-use characteristics, driver fixation points, and the road’s visual perspective. To evaluate consistency or discrepancies, the declared behaviours of survey respondents assessing these locations were compared with speed measurements collected from other drivers travelling through the same zones. The analyses help define the relationship between drivers’ perception and their actual behaviour, identifying which TCMs, when combined with specific road-environment features, are most effective in achieving the target speed of 50 km/h in built-up areas. The most effective chicanes proved to be those with the greatest width (2.5 m), i.e., almost equal to the width of a traffic lane, as well as those with a width of 2.0 m combined with a change in pavement surface from asphalt to stone paving, or those located upstream of a road section characterised by high curvature and limited visibility. In contrast, symmetrical islands, even with a width of 3.0 m, were found to be completely ineffective. The findings support the development of more effective transition-zone design principles and provide guidance for future mobility strategies, including the integration of automated vehicles in smart cities.

1. Introduction

As we can read in Ref. [1], the most recent WHO initiatives aimed at improving road traffic safety have had the desired effects and implementation of proven measures may substantially reduce road fatality rates. Currently, pedestrians, bicyclists, and motorcyclists account for over half of all road fatalities. As has been demonstrated in previous studies [2,3,4], the most dangerous areas are the speed-change locations in rural/built-up area transition zones due to an integrated and complex interaction of various components: driver psychology, traffic conditions, vehicle, surrounding environment, and existing road infrastructure. Different traffic safety analyses have shown the need for providing drivers with clear indications of changed surroundings and thus requesting speed reduction.

Various traffic calming measures (TCMs) have been implemented by road management authorities in different countries. The general definition of traffic calming was agreed upon at the 1996 Annual Meeting in Minneapolis, attended by traffic specialists from many countries [5]. The meeting was devoted to traffic arrangements and physical measures designed to control traffic speed and encourage driving behaviours adequate to given environmental conditions. For through-road sections running across villages and towns, the term transition zone (TZ) was defined in relation to traffic calming [6,7,8,9,10,11,12,13,14,15]. An illustrative definition of this term is provided in Figure 1.

According to [8,11], a transition zone is an area in which drivers should receive information about a change in the roadway environment (i.e., from a rural to a built-up area) and the need to adjust their speed. Importantly, it is not only necessary to reduce speed within the transition zone itself, but also to ensure that this reduction is maintained. Each transition-zone location is site-specific and requires its own unique solution. Consequently, measures applied in one transition zone may be more effective in some locations than in others. An important issue from the driver’s perspective is the creation of clearly defined rural and built-up areas, enabling drivers to recognise the point at which they are no longer travelling on a rural road segment and have entered a built-up area. Drivers in rural areas typically expect higher speeds and uninterrupted traffic flow, whereas drivers and pedestrians in built-up areas anticipate slower-moving traffic. To accommodate these differing needs and expectations, drivers should be made aware of the approaching built-up area and the need to reduce speed; therefore, there is a need for a transition zone prior to entering the built-up area. However, changes in geometric road elements require a gradual transition and should be appropriately communicated to drivers, including through modifications to the surrounding environment, for example, by means of noticeable and characteristic roadside features. Taking the above into account, a transition zone between these two environments may help drivers to appropriately adjust their speed and inform them about changes in the roadway environment. Nevertheless, it is important not only to reduce speed within the transition zone, but also to maintain this reduction within the built-up area.

The literature on traffic management distinguishes three zones at village or town entries, called the access, transition, and central zones, respectively [7,8,11,12,16,17,18,19]. Entering the transition zone and seeing D-42 and E-17 road signs, drivers are expected to slow down and abide by the applicable speed limit. Various TCMs can be used to achieve the desired slowing effect, including gateways, chicanes, horizontal deflections, and dynamic speed feedback signals (DSFSs) [6,7,8,11,12,20]. Different parameters and recommendations apply to different TCMs. Thus, the design of horizontal TCMs should consider the following parameters [6,7,8,11]:

-

Travel-path offset and angle [21] (p. 14—Figure 5);

-

Sight distance from the driver’s eye position onto the road section ahead and sight-line angular deflection [22] (p. 45—Figure 29 and p. 50—Figure 32);

-

Free-view width onto the traffic lane past the chicane [21] (p. 4—Figure 1) and [23] (p. 7—Figure 3).

The relevance of traffic arrangements and land development along the transition-zone sections characterising the local road traffic conditions has also been reported by researchers studying the subject [24,25,26,27,28,29]. It has been reported [30,31,32] that speed feedback signs or DSFS speed limit signs, the measures considered the option of choice in speed management applications, are effective only on a very short length of the road [13,24,25]. It is not uncommon to see drivers slow down right at the speed limit sign location, only to rapidly accelerate after passing it, the behaviour represented on speed profiles by the so-called kangaroo jump effect [32,33]. The so-called “kangaroo jump effect” describes a characteristic speed pattern in which drivers reduce their speed shortly before a TCM or DSFS and then rapidly accelerate after passing it [32].

Studying of traffic calming effects and the associated achievement of target speeds may concern different aspects and various speed parameters. The scope of such analyses may also be limited to the magnitude of speed reduction obtained by a given TCM or to speeds measured after passing the TCM [6,7,8,11,12,16,21,23]. The latter is the case in this study, which specifically deals with speeds after TCMs at town entrances (Figure 1). There are no universal guidelines for transition-zone length or placement of signs E-17 and D-42 (Figure 1), as these should take into account the local conditions, preferences of the local road authorities, and also town development plans [3,4,8,9,10,11,12,15,16,18,30,34,35,36,37].

Therefore, when investigating the effectiveness of a given TCM, the focus should be placed on the speeds measured at the transition-zone end, as it is the parameter representing the desired improvement in traffic safety and the well-being of the local community, and excessive speeds at this point were most likely the main reason for TCM application. Thus, well-supported, adequate siting of TCMs in transition zones (Figure 1) [6,8,11,17,24,38,39,40,41] and in the following parts of road sections within the town limits is considered essential for obtaining a sustained traffic-slowing effect [6,7,8,11,12,13,16,38]. Studies of existing TCM arrangements over transition-zone lengths have showed big variation in their siting [24,27,28,29,38,39] and a high variety of both transition-zone lengths [8,9,10,11] and the lengths of their effect on slowing the traffic [9,13,25]. The surrounding features [24,25,28,29,40,42] and road alignment, both vertical and horizontal [41,42,43,44,45,46], must also be taken into account when planning transition-zone and TCM arrangements to achieve the desired speed reduction.

Traffic simulators have been used in quite a number of studies devoted to speed reduction in transition zones [19,47,48,49,50,51,52]. The effect of gateways was also studied in this way [52,53,54,55]. In these studies, particular attention was paid to drivers’ fixation points, drivers’ visual attention, and the emotional mechanisms influencing visual attention under different cognitive tasks [22,53,56,57].

The growing importance of smart traffic management systems and cooperative driving for optimising road-space utilisation calls for studies that take into account the combined effect of various landscaping elements and applied TCMs on drivers’ perceptions [24,25,27,28,29,30,56,57,58]. Different speed management strategies and phasing-in of automated vehicles call for a joint use of speed feedback signs and D-42 (built-up area limit sign), additionally accompanied by appropriate surrounding features [24,25,27,30,58].

Talking about smart city design, it is worth mentioning the research of Lantieri et al. [24]. A traffic simulator was used in that research to investigate the effect of the road surroundings on psychomotor driving performance across twelve transition zones. The results showed that drivers are influenced by various elements present along the road, as well as by the details of TCMs. Eye movements caused by looking at roadside features (non-relevant driving targets) may result in distraction behaviour, and make drivers slow down as a consequence. Raised central islands were found to be the most effective measure to control distractions in gaze behaviour and cause speed reduction.

Based on the above review of previous research, and taking into account Lockwood’s definition of traffic calming [5], the objective of this study was to confront self-declared driving behaviours with free-flow speeds measured in selected transition zones. To this end, a survey was conducted in which respondents declared their reactions to traffic calming measures (TCMs) presented in photographs. They were not asked about the magnitude of speed reduction or the exact place where they would start braking, be it before or after the TCM in question, they only declared whether they would slow down or not.

Bearing in mind the contemporary challenges of digital traffic management planning, vehicle automation, and the smart city concept, this study focused on comparing declared behaviours with actual free-flow speeds, while taking into account several factors related to the surroundings of the transition zones.

This is different from the research to date, which has concentrated primarily on the speed reduction magnitude achieved by a given TCM, and on looking for the most effective TCM arrangements (including chicane type, design parameters, accompanying traffic signs posted in transition zone, gateway design parameters, etc.).

As demonstrated in the literature review, the issue of transition zones has been addressed in recent studies by numerous researchers; however, much remains to be studied. The results presented in this paper provide the next step in developing basic guidance for the design of high-to-low speed transition zones.

This approach is adequate when the purpose is to find an optimum TCM type and determine appropriate surrounding features. However, considering the present-day challenges related to smart city design and vehicular automation, the focus should be placed on the ultimate traffic calming effect, i.e., the final speeds when past a given TCM. Thus, the aim of this research was to determine which of the analysed TCMs, in combination with appropriate surrounding features, may lead to achieving the desired speed not exceeding 50 km/h, (speed limit for built-up areas in Poland). A research analyses sequence diagram is shown in Figure 2. The following research questions were formulated to address the above-mentioned issues:

➀

Is there a consistency in declared behaviours in the analysed transition zones in the respective age groups?

➁

Were the applied TCMs effective in lowering the speeds to the transition-zone speed limit in the analysed transition zones?

➂

Is there a consistency between the self-declared behaviours and the final speed past the TCM in the analysed transition zones?

➃

Which surrounding features turned out to be useful and facilitated attaining the target speed past the TCM in the analysed transition zones?

2. Research Assumptions and Method

2.1. Research Assumptions

The transition zones analysed in this study were chosen based on the following research assumptions. Forbes defined a transition zone as a section of road that is continuous with and connects a road section with a higher posted speed limit to a road section with a lower posted speed limit [8,11], and this definition was adopted for the purposes of this research. The transition-zone selection criterion was to obtain a sample differing in terms of the following:

-

TCM type: chicane, central island, refuge island, or DSFS;

-

Surrounding features: forest, fields, footpath(s), road or street cross-section, culvert or roadside ditch guardrails, bridge with parapets, etc.;

-

View of the road far ahead, including a road junction, close or distant buildings, no buildings in sight, refuge island, etc.;

-

Road alignment, both vertical (convex or concave arc) and horizontal (straight line, right-hand or left-hand curve).

In the selected transition zones, the following should be provided:

-

Chicanes on the entry lane (Figure 3a,b—2 m or 2.2 m wide), hereinafter referred to as chicane;

-

Chicanes with a central island separating the two traffic directions (Figure 3c,d—2 or 3 m wide), hereinafter referred to as central island;

-

Chicanes with a pedestrian crossing (Figure 3e—2.5 m wide), also hereinafter referred to as chicane;

-

Pedestrian refuge island (Figure 3f—2 m wide), hereinafter referred to as refuge island.

In Figure 3, in addition to the island width, the free-flow width a is also presented in order to better illustrate differences in its influence on driver perception.

The analysed cases should also include a combination of different TCMs within a single transition zone, such as a chicane combined with a change in the road surface texture, for example, from bituminous to natural stone block paving. Considering the research results presented in [24], all chicanes, central islands, and refuge islands should, if possible, be raised islands with a height of 14–16 cm.

All transition zones should be located on two-lane regional roads (voivodeship roads in the Polish administrative system) of up to 5000 veh./day average daily traffic. The length of the through road in a given town should be at least 0.3 km. The length of the transition zones should be approximately 100–200 m.

2.2. Transition-Zone Selection Process

Out of several dozen transition zones analysed, twenty-six were selected. They differed in terms of the following:

-

The applied traffic calming measures (TCMs);

-

Road surroundings and cross-section;

-

The view of the road ahead beyond the respective TCM, which included features within the driver’s field of sharp vision, such as a pedestrian refuge, an intersection, a bridge with railings, a culvert with road barriers, forested areas, and agricultural land;

-

Road profile (convex or concave curves) and horizontal geometry (straight sections, left-hand curves, or right-hand curves).

During the selection of transition zones, more than one hundred transition zones located on regional roads in the West Pomeranian Voivodeship were analysed. In the subsequent selection stage, several dozen transition zones were selected based on the criterion of similar traffic volumes. In the next stage, the criterion of good pavement condition was applied. Ultimately, 26 transition zones (Figure 4) were selected for the study, located at the entrances to small towns or villages with populations ranging from 150 to 2000 inhabitants. All but one of the transition zones ranged in length from 100 m to 200 m. The only exception was a 500 m long one, and this length was related to scattered buildings located further away from the road. All the transition zones were located on two-lane regional roads. The relevant speed limits were 90 km/h (approx. 56 mph) and 50 km/h (approx. 31 mph) in villages/towns. In one case there was a 40 km/h speed limit sign before the chicane due to a bridge with parapets located past the chicane on a left-hand curve of R = 165 m radius, which obscured the view.

2.3. Self-Declaration Survey

The survey questionnaire regarding driver behaviour in the selected transition zones was prepared on the basis of previous studies [25,51,59,60,61,62,63,64]. The questionnaire informed the respondents that the photographs show locations at entrances to small towns, either after or at the D-42 built-up area limit, with different surrounding features on both sides of the road and various TCMs. The respondents had little time to respond; this was in order to obtain their immediate reaction to the transition zone shown in the photograph. They were not given unlimited time to carefully analyse and reflect on each example. A total number of 102 respondents took part in the survey (

Table 1. Basic data about respondents. Source: author’s study.

Age Group	Age			Gender		Years Holding a Driving License
	Min	Average	Max	M	F	Min	Average	Max
Up to 30 years	20	23	28	22	10	2	5	10
From 30 to 50 years	34	42	50	14	18	8	21	33
Over 50 years	52	67	78	23	15	25	36	44

).

In addition to a photograph, each transition zone was characterised by a description of the traffic calming measures (TCMs), road surroundings, and cross-sectional and alignment characteristics. Each respondent was asked whether they would slow down if they found themselves in a given transition zone and could choose one of four responses: always, often, rarely, or never.

Based on the survey results, four study areas were identified, for which most of the respondents declared the following:

• A—slow down almost always (seven transition zones—from No. 1 to No. 7);
• B—often slow down (five transition zones—from No. 8 to No. 12);
• C—rarely slow down (six transition zones—from No. 13 to 18);
• D—never slow down (eight transition zones—from No. 19 to No. 26).

The subsequent comparative analyses of the survey results and traffic measurement data were conducted in the above-mentioned study areas.

2.4. Traffic Detection Method

Four traffic detection devices (SR4) were deployed on all the analysed transition zones. The detectors were time-synchronised, to measure speeds travelled by one vehicle at four locations. The SR4 sensors were mounted on existing signposts, lighting poles, and similar roadside infrastructure; where such elements were not available, additional poles were used. Consequently, their placement could exhibit minor variations in location (up to 5 m). Speeds were measured at four locations (Figure 5 and Figure 6):

-

SR4₁—15–20 m upstream of the beginning of the double solid line, depending on local constraints;

-

SR4₂—at the beginning of the double solid line, immediately before the taper of the hatched road markings leading to the island;

-

SR4₃—downstream of the taper of the hatched road markings, behind the island, at the end of the double solid line;

-

SR4₄—15–20 m downstream of the end of the double solid line, depending on local constraints.

According to the Polish Highway Code, the minimum length of a double solid line is 20 m. The SR4 deployment diagrams in transition zones including chicanes, central islands, or refuge islands are shown in Figure 5 and Figure 6. In transition zones including DSFSs, speeds were measured approx. 50 m before and approx. 15 m after the unit location (Figure 7 and Figure 8). For the sake of consistency of the speed data analysis in all transition zones, in relation to chicanes, central islands, and refuge islands, only the speed data from the first and the last speed detection devices were taken into account (Figure 6—before TCM—SR4₁ and after TCM—SR4₄).

The SR4 units simultaneously counted traffic and measured vehicle speeds in both directions. The following data were recorded: sampling time, vehicle speed, travel direction, sampling interval, sampling distance, vehicle length, and vehicle category. Speeds were measured with an accuracy of 1 km/h, and the sampling time was recorded with an accuracy of 0.1 s. The unit software enabled the identification of free-flow driving, defined as an empty lane ahead of the vehicle over at least the distance travelled at a given speed within 7 s [65]. The units were time-synchronised, meaning that identical measurement times were applied across all sensors. This allowed speeds and speed reductions to be analysed at the same time points. The survey took place in dry weather in summer. Traffic flow was smooth in both directions. Hourly traffic on the analysed sections varied between 150 and 450 veh./h. Considering the low traffic volumes on the analysed roads, the sensors were deployed for several hours. Based on preliminary measurements conducted in several transition zones and using standard statistical formulas (t_α = 1.096, α = 0.05, Δ = 0.05 v_av), the minimum required number of free-flow speed measurements was set at n_min = 100. In practice, however, after several hours of measurements, the number of analysed free-flow speed observations at each location always exceeded the minimum value n_min.

2.5. The ‘Before’ and ‘After’ Accident Rates

The accident rates before and after TCM implementation are typically considered in any traffic calming analyses. For the purposes of this research, these rates were taken from the SEWIK Polish registry of accidents and collisions [66]. Generally, the records covered a period extending a few years before and after the traffic calming project. In this period, over a dozen accidents occurred in some of the analysed villages/towns. In three villages, the accident rates concern the entire village and not only the analysed transition zones. The detailed accident data are given in Supplementary Materials.

2.6. Method Applied to Compare the Questionnaire Responses and Vehicle Speeds Measured Before and After TCM

All the speed data were subjected to standard statistical tests. All the speed datasets were normally distributed (Appendix A,

Table A1. Results of statistical tests. Source: author’s study.

	Equation (A1) 1		Equation (A2) 2	Equation (A3) 3
	Before TCM	After TCM
Transition zone No. 1	λ = 0.66	λ = 0.67	λ = 2.20	χ2 = 7.0
Transition zone No. 2	λ = 0.74	λ = 0.59	λ = 4.73	χ2 = 77.2
Transition zone No. 3	λ = 1.13	λ = 1.04	λ = 4.62	χ2 = 135.1
Transition zone No. 4	λ = 0.59	λ = 0.35	λ = 5.61	χ2 = 12.7
Transition zone No. 5	λ = 0.52	λ = 1.04	λ = 3.41	χ2 = 37.0
Transition zone No. 6	λ = 0.99	λ = 0.71	λ = 1.85	χ2 = 6.3
Transition zone No. 7	λ = 0.58	λ = 0.61	λ = 5.96	χ2 = 3.6
Transition zone No. 8	λ = 0.57	λ = 0.61	λ = 1.61	χ2 = 3.8
Transition zone No. 9	λ = 0.54	λ = 1.22	λ = 6.07	χ2 = 41.2
Transition zone No. 10	λ = 0.73	λ = 0.51	λ = 3.81	χ2 = 57.6
Transition zone No. 11	λ = 0.56	λ = 0.81	λ = 5.82	χ2 = 26.4
Transition zone No. 12	λ = 0.43	λ = 0.33	λ = 5.00	χ2 = 4.4
Transition zone No. 13	λ = 0.32	λ = 0.63	λ = 2.83	χ2 = 79.2
Transition zone No. 14	λ = 0.39	λ = 0.99	λ = 4.36	χ2 = 9.2
Transition zone No. 15	λ = 0.33	λ = 0.52	λ = 5.23	χ2 = 6.4
Transition zone No. 16	λ = 0.43	λ = 0.51	λ = 3.33	χ2 = 25.7
Transition zone No. 17	λ = 1.34	λ = 0.32	λ = 3.50	χ2 = 12.0
Transition zone No. 18	λ = 0.56	λ = 0.74	λ = 2.69	χ2 = 79.5
Transition zone No. 19	λ = 0.57	λ = 0.45	λ = 4.85	χ2 = 5.0
Transition zone No. 20	λ = 0.33	λ = 0.52	λ = 5.23	χ2 = 6.4
Transition zone No. 21	λ = 0.40	λ = 0.48	λ = 5.36	χ2 = 8.0
Transition zone No. 22	λ = 0.68	λ = 1.03	λ = 2.06	χ2 = 16.1
Transition zone No. 23	λ = 0.54	λ = 1.22	λ = 3.57	χ2 = 41.2
Transition zone No. 24	λ = 0.51	λ = 0.86	λ = 2.99	χ2 = 336.1
Transition zone No. 25	λ = 0.54	λ = 1.01	λ = 1.41	χ2 = 3.6
Transition zone No. 26	λ = 0.48	λ = 0.82	λ = 1.90	χ2 = 15.4

¹ Kolmogorov–Smirnov goodness-of-fit test, λ_α = 1.36, α = 0.05. ² Two-sample Kolmogorov–Smirnov test, λ_α = 1.36, α = 0.05. ³ Median test, χ_α² = 3.84, α = 0.05.

). The questionnaire responses and the measured speed data are presented as percentage bar charts, giving percentages of the respective ranges. In addition, the speed data are presented in box plots. The results of the statistical tests comparing the speed populations before and after the TCM are presented in Appendix A.

3. Study Areas

The selected transition zones are presented in detail in the Supplementary Materials attached to this article, which include photographs, questionnaire responses, transition zone descriptions, speed survey data, TCM parameters, hourly traffic, and accident numbers. Photographs of the analysed study areas are presented in Appendix B.

4. Results

The questionnaire responses and speed survey results in the selected study areas are presented as percentage bar charts. The charts in Section 4 show, in addition, the horizontal TCM type: chicane with island on entry lane, central island, chicane with a pedestrian crossing, refuge island, and chicane accompanied by a change in surface texture. DSFSs are also indicated by acronym. The whiskers in the box plots represent the minimum and maximum values, the lower and upper edges of the boxes determine the first and third quartiles, and the bold line designates the median value (v₅₀).

For the sake of clarity, the box plots in Section 4 also show the speed limit of 50 km/h marked with a continuous red line and B-33 sign (speed limit 50 km/h). Considering the interpretation of the speed limit given on the B-33 speed limit sign, which in some countries may be exceeded, yet to no more than 59 km/h, the latter value was marked on the charts with a dashed line (50 + 9 km/h). This is in line with the Convention on Road Traffic definition [67].

To address research question ①, each transition zone is attached with questionnaire results by age group in the Supplementary Materials. The charts give the percentages of respondents who declared that they would slow in a given transition zone. In the case of respondents declaring ‘never slow down’, only the corresponding percentage share is given.

4.1. Results in Study Area A

The questionnaire responses and speed survey results in study area A are given in Figure 9. In study area A, between 75% and 87% respondents responded that they would almost always slow down in a given transition zone (Figure 9a). The survey did not ask about the slowing magnitude or whether the driver would start slowing on the approach to or when passing a given TCM. The questionnaire asked solely about the driver’s behaviour. Analysis of the data presented in the chart (Figure 9a) showed that the type of TCM was not the primary factor influencing the respondents’ behaviour. A detailed analysis of the transition zones in the study area A examples shows that the respondents took into account the street-type cross-section details and surrounding features, including footpaths, barriers, and houses in close vicinity. However, the self-declared behaviours were not fully consistent with the speed survey results (Figure 9c,d). Reductions in ∆v₈₅ and ∆v_av were noted in the analysed transition zones (Figure 9d); however, only in the following three cases was v_85 after approximately 50 km/h:

-

No. 2—chicane with a pedestrian crossing (2.5 m wide island) with buildings in close vicinity;

-

No. 3—chicane (2 m wide island) with a change in surface texture to stone block paving;

-

No. 5—chicane (2 m wide island), located before a very curvy street section, enhanced by culvert guardrails in the driver’s central vision area.

In the other cases (Figure 9d—Nos. 1, 6, and 7), despite considerable slowing, the final speed v_{85 after} still exceeded 50 km/h. An increase in the speed was also noted in one case (Figure 9d—No. 4 and Appendix B—Figure A1) where there was a common transition zone for two neighbouring towns. The above observations are in line with the findings of other researchers [8,11,24,25], and thus confirm the recommendations formulated by these research teams regarding the need for a thorough analysis of the surrounding land use and features of a transition zone before determining the location of a specific TCM, and, if necessary, the implementation of additional traffic calming treatments. The latter may include delineator posts with retroreflective elements or LED lights, changes in surface texture, roadside vegetation, optical speed bars, or rumble strips, as well as the use of additional traffic islands—both larger elements such as build-outs and smaller pinch points, hereinafter collectively referred to as traffic islands [8,9,11,24,26,29,39,45]. Traffic islands used as supplementary treatments were found to considerably contribute to speed reduction due to horizontal deflection of the driving path and sharp hatched-area taper rates (1 in 5 ≈ 11.3° or 1 in 7 ≈ 8.1°) [24,26]. Apparently, traffic islands placed before a chicane or a refuge island could effectively reduce the final v₈₅ speed to the desired level of ≤50 km/h.

The questionnaire responses for study area A were analysed to check their applicability to smart city design, including intelligent traffic management and planning of transition-zone traffic calming schemes. This analysis showed consistency of responses in all age groups in the case of transition zones, including various road infrastructure components (Figure 9—Nos. 1, 2, and 3). In the other cases under analysis (Figure 9—Nos. 4, 5, 6, and 7), identical behaviours were declared by the 50+ age group members, irrespective of the surrounding features, road geometry, and view down the road past the TCM. In the other age groups, between 72% and 82% slowed down in the case of a street-type cross-section and between 63% and 72% in the case of a road-type cross-section. These results indicate the need to use additional road infrastructure elements in order to effectively lower the speed at a smart city entrance to below 50 km/h. Automated vehicles would require speed feedback signs of 50 km/h placed under the D-42 traffic sign.

Automated vehicles are an important part of future mobility solutions. Other than the drivers, whose behaviours are subjective, often different from their former self-declarations, automated vehicles reactions to road infrastructure are governed by sensors, digital maps, and control routines [68,69,70]. When there is only one horizontal TCM at the town entrance, such as a chicane or refuge island, without an accompanying D-42 sign, an automated vehicle will react to the horizontal deflection, but it will not necessarily slow down to 50 km/h, unless guided to by speed feedback signs, HD maps (high-definition digital maps covering the area of roads and roadside features, including traffic lanes, speed feedback signs, and road infrastructure), or a V2X communication system (allowing vehicles to interact with their environment using advanced software and 5G technology) [68,69,70]. Thus, in years to come, the transition-zone effectiveness will depend even more on the interaction of physical and digital road infrastructure components, comprising conventional traffic signs and automated vehicle-compatible elements. Thus, the analysis of the study area A results showed that for effective speed reduction, conventional TCMs must be used in combination with supplementary physical and digital infrastructure elements, which will have the same effect on drivers and on vehicular automation systems.

4.2. Results in Study Area B

The questionnaire responses in study area B are very similar, irrespective of the chicane type, cross-section type, surrounding features, and road alignment (Figure 10a). Also the final speeds past the chicane were similar (Figure 10b). In study area B, speed reduction was noted in all the analysed transition zones (Figure 10d). This was, most likely, related to the surrounding features, limited view of the road far ahead, and last, but most importantly, road alignment. The analysis of the data presented in Figure 10c showed very small speed variations in transition zones No. 10 and No. 12. Although speed reductions were noted in these transition zones, the speed ranges varied very little, undoubtedly due to the cross-section and, even more importantly, the surrounding features. In the next transition zones, speeds were much more varied and speed reductions were closer (Figure 10d), most likely due to there no longer being a good view of the road far ahead and the view’s alignment, both within a forest and across fields. The analyses in study area B led to an important conclusion: that variations in speed and similar reduction amounts (Figure 10c,d) were not at all related to the chicane type. Similar to study area A, in study area B the questionnaire results were also analysed to check their applicability to smart city design, including intelligent traffic management and planning of transition-zone traffic calming schemes. This analysis showed consistency of responses in all age groups (from 59% to 69%) for all but one of the analysed transition zones (Figure 10b). A smaller percentage of self-declared slowing was noted only for transition zone No. 8 in the below 30 age group (53%). A comprehensive analysis showed that, with 90% slowing declarations, the respondents aged 30 to 50 were the most disciplined of all age groups. However, this situation was limited to roads with street-type cross-sections or when the view down the road was obscured by a horizontal or vertical curve in the road alignment past the chicane. This result is almost fully consistent with the findings of Hallmark et al. [46], that while open spaces offer drivers fast mobility, in built-up areas they have to take into account the presence of pedestrians and cyclists, typical of town environments. Driving through transition zones with poor road infrastructure, the drivers passing through a village without the intent to stop may not receive sufficient visual indications of changed surroundings and may not have sufficient time to respond and adjust their driving speed. These results indicate the need for additional road infrastructure elements in order to make drivers slow down to below 50 km/h at a smart city entrance.

The 50 km/h speed feedback signs must be placed to guide automated vehicles. Roadside vegetation may be preferred over delineators in transition zones where farm vehicles (carrying unfolded attachments) and HGVs are a consideration [8,11,24,25,27,28,29].

4.3. Results in Study Area C

In study area C, the ‘rarely slow down’ option was chosen by between 46% and 71% of the respondents (Figure 11a).

The TCM used in the transition zone in question (such as a central island 3 m wide, refuge island, chicane with island on the entry lane, or DSFS) was not the primary factor influencing drivers’ self-declared behaviour. Also, the cross-section type, surrounding features, and road alignment appeared to have no significant influence in this respect. However, these responses were not consistent with the speed survey results (Figure 11). The analysis of the data presented in Figure 11d showed that the drivers slowed down in all the transition zones in question, yet never to less than 50 km/h. Only in three transition zones was the value of v_{85 after} very close to 60 km/h (Figure 11d—Nos. 15, 16, and 18). It is most probably the roadside elements, including road verge delineators at pedestrian refuges or chicanes or white painted boulders that actually influenced the drivers’ behaviours. These elements were positioned in the driver’s central vision area.

Also, in study area C the questionnaire results were analysed to check their applicability to smart city design, including intelligent traffic management and planning of transition-zone traffic calming. Slowing was self-declared by similar numbers of respondents, except for in the transition zone in Figure 11—No. 14 (which included a refuge island located past the chicane)—and the transition zone in Figure 11—No. 13 (DSFS, featuring a right-hand curve and obscured view past the curve)—where only 4–7% of the respondents did not firmly declare slowing. All these transition zones had a road-type cross-section, giving the participant not enough indications of changed road surroundings, in line with the findings of Hallmark et al. [46]. Therefore, similarly to study area B, and in line with the conclusions of previous research projects [8,11,24,25,28,29], the authors’ applied additional road infrastructure elements in all the transition zones in study area C.

4.4. Results in Study Area D

In study area D the respondents firmly declared that they would not slow in the transition zones presented to them. Their choices were, most likely, based on very good visibility of the road far ahead on straight sections of the road with road-type cross-sections, both in forests and across fields. Only in the first two transition zones was there a horizontal curve in view, at a distance of approx. 300 m past the chicane, and this was probably the reason for there only being 36% to 42% ‘never slow down’ responses (Figure 12a,b—Nos. 19 and 20). In the other transition zones, between 54% and 78% declared no slowing in any road surroundings, and this despite a village skyline and houses visible at a distance, the relevant traffic signs, and a TCM in place (Figure 12a,b—Nos. 21–25). The last transition zone cut across a rural area, and with a poorly visible village skyline at about 1 km and no obstacles ahead (Figure 12a,b—No. 26), the ‘never slow down’ option was chosen by 84% of the respondents. This high percentage indicates the ineffectiveness of the existing central island, deflecting the path by only 1 m, and could give grounds for questioning the use of TCMs in general. In this case the central island preceded the built-up area limit sign by approx. 150 m (Figure 12a,b—No. 26). On the right-hand side, the town-entry lane was bordered by farmland with no buildings, exits, etc., which makes the existing narrow central island a solution that is not supported by any traffic calming guidance.

5. Discussion

5.1. Correlation of the Survey Responses and the Traffic Survey Speed Data

Turning to research question ③, we analysed the correlations between the survey responses and speed survey data. This covered analysing the percentages of ‘always’ and ‘never’ responses and initial (‘before TCM’) and final (‘after TCM’) v₈₅ and v_av speeds. The relevant values of the R correlation coefficient are given in

Table 2. R correlation coefficients. Source: author’s study.

Study Area	R = f (Percentage of ‘Always Slow Down’ Respondents; v)				R = f (Percentage of ‘Never Slow Down’ Respondents; v)
	v85 before	v85 after	vav before	vav after	v85 before	v85 after	vav before	vav after
Study area A	−0.11	−0.28	−0.07	−0.21	−0.13	0.08	−0.20	0.06
Study area B	−0.16	0.11	−0.48	0.02	0.54	0.50	0.25	0.33
Study area C	−0.47	−0.54	−0.30	−0.55	−0.35	−0.34	−0.65	−0.36
Study area D	0.72	0.06	0.69	−0.01	−0.73	−0.15	−077	−0.15

.

Analysis of the data presented in Table 2 showed that the highest values were obtained in study area D, and the regression graphs in Figure 11 concern these cases. The calculated correlation coefficients show that in study areas A, B, and C, the values of the v₈₅ and v_av ‘before’ and ‘after’ speeds decreased with the increasing percentage of ‘always slow down’ responses in almost all cases, as indicated by their negative values. In Table 2, positive values of the correlation coefficient R obtained for study area D are highlighted in bold, as only in this area was it demonstrated that as the percentage share of respondents declaring “always slow down” increased from 0% to 23%, the analysed speeds (v₈₅ and v_av) also increased. Conversely, when respondents declared that they “never slow down”, the speeds decreased. These differences can be explained by the fact that speeds were measured in situ, and in this case different drivers made different decisions under real traffic conditions, dynamically analysing the changing roadway environment and traffic situation. In contrast, respondents made their decisions based on static images and an empty road, without the presence of other vehicles. Photographs of the transition zones analysed in study area D are presented in Appendix B in Figure A4 and in the Supplementary Materials. To illustrate the basis of the observed differences, the numbers of the analysed transition zones are additionally indicated in Figure 13.

5.2. Comparison of the Survey Responses with the Speed Survey Data, Taking into Account Driver Fixation Points and Visual Attention Areas

As mentioned, the analysis of driver perception and fixation points is an important part of TCM planning in transition zones [22,25,53,56,57,59,61,64,71]. Studies on driver fixation points (Figure 14) and central vision area determination in relation to speed (Figure 15 and Figure 16) date back to the 1960s [72,73]. These issues have been the subject of many research projects related to the development of road design guidelines and determination of their main parameters.

Designation: pink dots—eye fixation points; percent values—attention allocation on different road elements.

In order to answer research question ④, three visual attention regions were distinguished in the driver’s field of vision, taking into account good sighting, eyeball movement, and head turning to assess road situation (Figure 15) [25,48,56,63]. The central vision area varies with driving speed (Figure 15, indicated by dashed lines). Studies on fixation points [25,73,74,75,76,77,78] allow the determination of drivers’ focus areas depending on driving speed (Figure 16).

The analysis of these areas and fixation points (Figure 14 and Figure 16) gives guidance on which elements of the road and its surroundings should be considered particularly relevant in the smart city transition-zone design. Particular attention should be paid to raised islands of chicanes and refuge islands, to which drivers allocate approx. 43.5% of their total attention, as reported by Lantieri et al. [24]. Also, hatched areas on the way to TCMs are very important, as they receive approx. 28.6% of drivers’ attention [25,72,73]. In the analysed transition zones, the hatched areas before chicane islands or central islands had taper rates of 1 in 10 (approx. 5.7%) or 1 in 15 (approx. 3.8%). Hatched areas with 1 in 5 (approx. 11.3°) tapers were limited to refuge islands. The results of studies on the effect of pavement markings on the entry speed [8,24,25,79] showed equal importance of transverse markings, i.e., rumble strip or dragon’s teeth.

This is probably related to fixation points located at the road centreline, part of the opposing traffic lane, and the outer roadway edge (Figure 16; 6.9% and 2.4%, respectively), where pavement markings are typically placed. The next fixation-point total concerns road shoulders, receiving 7.3% of drivers’ attention. Therefore, delineators are installed or vegetation is planted on the shoulders of roads with a rural-type cross-section or on footpaths along roads with a street-type cross-section [8,11,24,25,28,29]. However, the effect of roadside vegetation alone, as indicated by the studies in [24,25,80,81,82,83], may prove insufficient if it is not combined with appropriate physical interventions.

Based on the analyses conducted and the need to address research question ④, the authors emphasise that particular attention should be paid, when planning transition zones in rural areas, to the design of the roadside, which receives approximately 7.3% of the driver’s attention. During the in situ studies, it was found that approximately 5–8% of passenger car drivers on free-flowing roads bypassed pedestrian refuges, central islands, or chicanes on the left, incorrectly driving in the lane designated for opposing traffic. In such cases, speeds significantly exceeded the permissible limit (measured speeds were 80–100 km/h or higher). Edge posts or roadside vegetation implemented along the shoulder considerably reduce the occurrence of such improper manoeuvres [11,24,25,28].

However, in agricultural areas, the main issue is the passage of farm machinery, particularly grain combines, as farmers often do not fold the header and may catch it on the delineators [28,29].

An example is the chicane in transition zone No. 3 (Figure 17), where, during the road reconstruction, delineators were installed on both the left and right sides of the road. The delineators on the right side were primarily intended to enhance safety and improve the visibility of the road curvature before and after the chicane, while the two delineators on the left side were mainly intended to prevent improper manoeuvres at the entrance to the village in the lane of opposing traffic. However, after a few years, the delineators were no longer present, as combine harvesters driving with the header unfolded caught on them, and they were gradually destroyed. The worn shoulder serves as a silent witness to vehicles travelling with incorrect trajectories in the absence of delineators (Figure 17b). The width of the deformed shoulder increased from 0.30 m in 2017 to 1.2 m in 2025. During speed measurements conducted in 2025, all heavy vehicles, farm machinery, combine harvesters, and delivery vans travelled with their outer wheels on the unpaved shoulder. This type of in situ driving was also observed for most passenger cars when drivers passed along the chicane at speeds exceeding the 50 km/h limit (Figure 18). In measurements conducted in 2017, when delineators were installed, such cases were rare.

5.3. Analysis of Speed Data Depending on TCM Type

The recommendations regarding transition zones’ surrounding features given in [8] (p. 79—Figures 4–19), and refs. [27,29,80,81,82,83] were used as the basis for the study of the influence of different TCMs present in the analysed transition zones (central island, chicane, raised island, and DSFS). As an answer to research question ②, free-flow v₈₅ and v_av ‘after’ speeds were grouped by TCM type and are shown in Figure 19. The speeds noted in transition zones including central islands and refuge islands were much above the speed limit, regardless of the surrounding features, view of the road head past the island, or road alignment. Lower speeds were noted in transition zones No. 7 and No. 8 that had street-type cross-sections and included refuge islands: v_av of slightly above 50 km/h, and v₈₅ of over 60 km/h (Figure 19). This indicates the ineffectiveness of central islands, even when 3 m wide and preceded by a very sharp hatched area of 1 in 5 taper (Figure 19—No. 14). For effective traffic calming, these transition zones would require placement of bollards on footpaths or delineators with LED lights on shoulders, depending on the road cross, and also roadway narrowing or additional traffic islands (Figure 20) [11,24,25,26,27,28,29].

Among the transition zones from chicanes, the ‘after’ free-flow speeds v₈₅ and v_av fell below the speed limit only in (Figure 19):

-

Transition zone No. 2, including a 2.5 m wide chicane with a pedestrian crossing before a bridge, and with buildings in close vicinity;

-

Transition zone No. 3—2 m wide chicane with pavement change from asphalt to cobblestone;

-

Transition zone No. 5, including a 2 m wide chicane before a culvert lined with visible guardrails and a very curvy road section past the chicane.

In the other transition zones with chicanes, the free-flow speeds exceeded the speed limit of 50 km/h regardless of the surrounding features, view of the road ahead, or road alignment. Therefore, in line with the conclusions of previous research [8,11,24,25,28,29], the recommended treatments include the use of delineators (Figure 17a) or roadside vegetation planted along a chicane, a central island, a refuge island, or a traffic island located in front of them (Figure 18). Hatched-area tapers of 1 in 5 or 1 in 10 before 2 m wide islands will not suffice to achieve speed reduction to 50 km/h. Sharp tapers may lead to potentially dangerous driving over the island corners (Figure 21).

5.4. Proposed Sequence of Analyses in the Transition-Zone Design Process

Considering different transition-zone preferences for autonomous and conventional vehicles, alternative TCM solutions are proposed in Figure 22. Speed feedback signs set at 50 km/h will cause any autonomous vehicle to slow down and comply with the speed limit while crossing the town. This is related to the fact that autonomous vehicle software is programmed to adapt to information provided by road signs, in this case a speed limit of 50 km/h (Figure 22, left-hand side). In contrast, according to the results of the authors’ research and studies by other researchers cited in this paper, drivers of conventional vehicles should be influenced in transition zones by additional factors that more dynamically affect perception and the expected speed reduction. In this case, these include the following measures (Figure 22, right-hand side): shrubs along road shoulders, traffic islands with planted greenery and raised kerbs, bollards, optical speed bars or rumble strips, and dragon’s teeth.

Given the distant location of surrounding houses and the straight road alignment, a chicane with an island at least 2 m wide should be placed on the entry lane. When designing for current conditions while considering future smart city entrance requirements, the chicane width should be increased to at least 3–5 m, depending on available land, and roadside vegetation should be provided along the chicane and on the chicane island as an additional treatment. The benefits of such vegetation have been confirmed in previous studies [8,11,21,52,55,80,81,82,83]. Constructing a chicane with an island wider than 2 m entails higher costs due to additional land acquisition (if available) and modification of roadside ditches. Roadside vegetation is a more cost-effective solution, even if periodically damaged by farm vehicles, requiring replacement with new shrubs. Shrubs planted along the road shoulder have the greatest impact on passenger car drivers’ perception by visually narrowing the lane and posing a potential hazard for vehicle body damage when drivers follow incorrect trajectories or exceed the 50 km/h limit [21,28,29,52,55,80,81,82,83]. Furthermore, gravel shoulders are not recommended in rural areas, as they may encourage drivers to use the shoulder and exceed the 50 km/h speed limit. If post-construction surveys indicate speeding, additional treatments are recommended, including DSFS (Figure 22) combined with transverse pavement markings (optical speed bars, rumble strips, or dragon’s teeth) [8,11,24,25,40,79], or traffic islands (build-outs or pinch points) [26,27,28,29,39].

6. Conclusions

6.1. Limitations

The analysed transition zones were all located on low-volume regional roads. They included chicanes, central islands, refuge islands, and DSFSs, but extrapolation to other regions and traffic conditions is limited due to a lack of gateways, rumble strips, or other physical measures of this kind.

The self-declared driving behaviours were based mainly on the photographs shown to the respondents, which did not truly represent the dynamic perception conditions in a live traffic environment. Also they did not consider behavioural changes over time, thus disabling a longitudinal analysis.

6.2. Summary

Respondents’ declarations differed between age groups and study areas and did not always correlate with the actual measured speeds. Older respondents were more disciplined and reported slowing down in the transition zones of study areas A and B across various TCMs. The other two age groups reported similar behaviours in these transition zones, but with greater variability. In contrast, in study areas C and D, declarations across all age groups were highly varied, regardless of the type of TCM. A surprising survey result was that most respondents reported rarely slowing down before pedestrian refuges or a 3 m wide central island, which may indicate the influence of surrounding land use on their behaviour and a completely inappropriate selection of TCM types for the given conditions. This highlights the need for further research under dynamic conditions using methods such as eye-tracking or driving simulators.

Individual TCMs rarely resulted in speed reductions below 50 km/h. Their effectiveness increased only when combined with other elements of surrounding land use, limited road sight-lines, and additional infrastructural measures. Particularly promising is the use of lateral islands as traffic islands, which force drivers to adjust their horizontal trajectory, thereby increasing cognitive demand and, consequently, achieving effective speed reduction.

The results demonstrate a clear synergistic effect between traffic calming measures and the surrounding environment. Chicanes combined with street-type cross-sections, sidewalks, nearby buildings, or limited sight distance resulted in post-TCM free-flow speeds closest to the target value of 50 km/h. In contrast, similar measures implemented in open rural environments typically produced post-TCM speeds exceeding 60 km/h, despite frequent driver declarations of slowing down. This confirms that the environmental and visual context plays a decisive role in achieving effective speed reduction in transition zones.

In many cases, there was a discrepancy between declared behaviour and actual free-flow speeds. Drivers reported slowing down more frequently than was observed in the measurements, yet the free-flow speed after TCMs was very rarely below the limit (50 km/h). This confirms that subjective perceptions cannot replace empirical data, and that the assessment of TCM effectiveness must be based on field measurements.

The design of the road surroundings has a significant impact on drivers’ perception and the effectiveness of TCMs. Strategically placed vegetation along the roadway creates an optical narrowing effect, which encourages speed reduction. Both the amount of vegetation and its distance from the edge of the carriageway are important. These findings are consistent with those of other researchers, indicating that an appropriate combination of infrastructural and landscape elements enhances the effectiveness of transition zones.

Since appropriately designed transition zones may contribute to traffic safety improvement by effectively reducing the traffic speed, this research offers practical insights for policymakers, engineers, and mobility planners, supporting their further actions in the context of smart city planning and infrastructure design.