Roundabouts and other intersections

An intersection is a location where roads and traffic flows cross one another or split and where road users may change direction. Intersections come in different shapes and sizes, but are characterised by the fact that road users cross one another at the same level. A roundabout is an at-grade intersection at which traffic flows in a circular movement and where traffic already on the roundabout has right of way [1]. A comprehensive description of different intersection types can be found at the section What are the different types of intersection? An intersection differs from a crossing. At a level crossing different traffic flows meet, but road users may not change direction. Examples are railway crossings or pedestrian and bicycle crossings. At split-level crossings, road users do not meet.

Intersections may be subdivided by type, priority regulation and layout.

Intersection types

Table 1 shows the different intersection types.

Table 1. Intersections by type[i] . *In the figure GOW stand for distributor road and ETW for access road.

About one third of all road deaths can be attributed to intersections[i]. The share of road deaths on urban intersections (almost half) is higher than that on intersections outside the urban area (somewhat under a quarter). Among cyclists, the number of road deaths on intersections is higher than on road sections: an annual average of 54% of the total number. A second noteworthy category are (light) moped riders: half their annual number of road deaths can be attributed to intersections. For the other modes of transport, the share of fatal intersection crashes is much smaller when compared to the number of fatal road section crashes. Examples are fatal car crashes at intersections at 19% and pedestrian fatalities at 26% (see table 2 below). Table 2 shows the 2015-2019 average number of road deaths at intersections compared to the number at road sections, differentiated for urban or non-urban area, and for mode of transport. The table shows absolute numbers and percentages.

Table 2. The average annual number of BRON-registered road deaths at intersections and at road sections in 2015-2019, disaggregated for transport mode and location in or outside the urban area. Of 7% of fatal crashes at intersections it is unknown whether they occurred in or outside the urban area. Moreover, an actual 15% of road deaths remain unregistered. For these, location data are unavailable, not even whether the crash occurred at an intersection or road section[ii].

According to Sustainable Safety, the ideal intersection depends on the road categories of the intersecting roads. For more information about Sustainable Safety, see SWOV fact sheet Sustainable Road Safety. Table 3 indicates which intersection is recommended at which location. The recommended intersections are presented in order of safety, from most recommended to least recommended. An additional note: At 50-50/50-70/70-70 (distributor-distributor road) intersections, safety depends on the capacity of the intersection. Intersection capacity in its turn depends on the number of lanes on the legs and the traffic distribution across the legs. The table is based on the assumption that each intersection has the same traffic volume. However, because a priority intersection has a lower capacity, it cannot cope well with a large amount of traffic when demand is high. For busy intersections, a signalised priority intersection is therefore safer. The section under the question Which intersection types are safest? explains why one intersection type is safer than the other.

Table 3. Recommended intersection types. The safest and therefore preferred type is mentioned first.*In the figure GOW stand for distributor road and ETW for access road.

The guiding principle for sustainably safe intersections is that conflicts between vehicles may and can only occur when vehicle speed and mass do not strongly diverge [9]. For the sake of road safety, traffic calming measures near or on intersections are therefore required. At intersections with distributor roads in the urban area, where motorised traffic meets, a maximum speed of 50 km/h is considered safe. In situations where heavy motorised traffic intersects with the much lighter (light) mopeds and bicycles or with pedestrians, speed should be reduced even more. For example, for intersections with distributor roads in the urban area where pedestrians and cyclists intersect with motorised traffic, the safe maximum speed is 30km/h [9]. Lower speeds can be enforced by roundabouts, speed humps just before the intersection, or a raised junction. In addition, sustainably safe intersection requirements are intended to exclude certain combinations of crash opponents (for example trucks and bicycles) being involved in conflicts in certain circumstances, since the implemented traffic facility virtually excludes these encounters [9]. For the purpose of sustainable safety, the following intersection facilities are therefore recommended (Table 4):

Table 4. Main intersection facilities to prevent common conflicts or collisions. For a more detailed description of the conflict types, see the section What characteristics affect intersection safety? And for information on deflection of the bicycle track and the bike box, see the section How to make intersections safer for cyclists?.

Research [9] shows that roundabouts are the safest intersections. The list of intersection types below is in order of safety, from safest to least safe[i]:

1. Roundabout. According to research comparing the crash risks of different types intersections [9], roundabouts are safer than conventional intersections. A recent publication [10], however, came to the contrary conclusion that more traffic injuries occur at roundabouts than conventional intersections. It is not clear to what degree this recent result is influenced by the fact that relatively unsafe intersections are more likely to be converted into a roundabout than relatively safe intersections. Before-and-after studies conducted at the same locations [11] [12], have shown that the conversion of an intersection into a roundabout lead to a substantial reduction in traffic casualties.
2. Prioritised intersection. Traffic volumes at these intersections are often lower than at signalised intersections. The design of a priority intersection is simpler than that of a signalised intersection. This results in less complex situations and, thus, fewer crashes.
3. Signalised intersection. These intersections are generally less safe than other types of intersection. The choice for a signalised intersection is often made when other intersection types would not be able to handle the high number of passing motor vehicles. Signalised intersections are often complex because of their higher capacity, which may reduce their safety. The capacity of an intersection is determined by the number of lanes of the legs and by traffic distribution across these legs.

Furthermore, a new intersection has recently been introduced: the bow-legged intersection (for a description, see the section What are the different types of intersection? ). It is, however, still too early to include the bow-legged intersection in the list above, since bow-legged intersections have not yet sufficiently been evaluated. There are, however, some graduate studies of bow-legged intersections. They generally conclude that a bow-legged intersection has more points of conflict than a single-lane roundabout. The points of conflict are, however, further apart than at priority intersections, the types of conflict are different than at other intersection types and conflicts occur at lower speeds.

For more information about the safety of bow-legged intersections, see the following three graduate studies: [13] [14] [15].

There are no legal requirements for the design of different types of intersection. Sustainable Safety does, however, include recommendations (for a description, see the section What is the ideal intersection according to Sustainable Safety? ) and CROW also makes several design recommendations:

• Basic characteristics intersections and roundabouts (CROW) – Design recommendations for intersections in and outside the urban area [1]
• Road Design Manual (CROW) – Design recommendation for intersections in and outside the urban area [8]
• ASVV (CROW) - Design recommendations for intersections in the urban area [7]

Since there are no legal requirements for the layout of intersections (nor the layout of road sections), uniformity is not guaranteed. Guidelines are easier to circumvent than legal requirements. This is as it should be, since the situation is different everywhere and therefore a similar design for every intersection would be impossible. Furthermore, the space available for the intersection and the environmental characteristics also come into play. They affect what an intersection will look like in a given situation.

The first determining characteristic for intersection safety is the type of intersection. Signalised intersections, for example, are less safe than other intersection types because of their capacity, high traffic volumes and complexity (also see the section Which intersection types are safest?)

Moreover, each intersection type has characteristics that may affect its safety:

• the number of possible points of conflict;
• speed;
• differences in mass of vehicles that meet at the intersection;
• impact angle;
• the amount of traffic, broken down by the number of passing motor vehicles and the ratio between main flow and lateral flow.

Intersection safety strongly depends on the number of possible points of conflict: the fewer, the safer. A conflict between vehicles should only be possible when speed and mass differences are small. The impact angle is another important factor. The angle is determined by the type of conflict. Table 5 shows four different conflict groups.

Table 5. Four conflict groups [9].

A last important characteristic is the traffic volume: the more passing vehicles, the more crashes. The way in which this contributes to intersection safety was studied in two ways. For the urban area, the number of passing vehicles was studied [9]. This is the sum of lateral flows and main flow. For intersections outside the urban area, the ratio between lateral-flow volume and main-flow volume was studied. This ratio is strongly related to the number of injury crashes [16][i].

Before-and-after studies show that intersections become safer after implementation of traffic lights [17] [18]. However, when the number of injury crashes at signalised intersections is compared to this number at other intersection types with similar traffic volumes, signalised intersections are the least safe intersection type. Unsignalised intersections and roundabouts have fewer injury crashes [9] [19]. So, after adjusting for traffic flow to determine risk, signalised intersections are still less safe.

In addition: traffic lights are often used when the number of passing motor vehicles is too high for other intersection types to cope with. They enhance traffic flow.

Roundabouts are the safest intersection type because of the following three characteristics:

• They have fewer points of conflict (see Figure 1);
• Speed is reduced, especially when the legs have perpendicular connections to the roundabout;
• The impact angle is smaller.

In terms of road safety, roundabouts should therefore be preferred to other types of intersection. First, single-lane, four-legged roundabouts only have 4 conflict points, compared to the 24 conflict points of ordinary four-legged intersections [9]. This difference is easy to spot when looking at Figure 1 below. The median island ensures fewer conflicts, since motorised traffic can only flow in one direction. This makes a roundabout less complex than other types of intersection.

Figure 1. The differences in conflict points between roundabouts, three-legged and four-legged intersections.

Second, because of the layout and priority regulation of roundabouts, all traffic flows need to reduce speed, whereas this is not the case at many other types of intersection [20]. The median island ensures that road users cannot cross in a straight line. Together with the perpendicular location of the legs, this has a speed reducing effect. There are roundabouts with deflecting legs or multiple lanes; they have a lesser speed reducing effect. The priority regulation reduces speed since road users intending to enter the roundabout have to give right of way to traffic already on the roundabout. Third, the layout of a roundabout makes for an entry angle that is smaller than that of other intersection types, which also diminishes the impact angle. Certain conflict types therefore no longer occur on roundabouts [21]. On a roundabout, there are more lateral conflicts which are relatively less serious, whereas at other intersection types, frontal conflicts, that are often relatively serious, may also occur.

The list below describes the safety of single-lane roundabouts, multiple-lane roundabouts, and turbo roundabouts:

• Single-lane roundabout. Four-legged, single-lane roundabouts are safer than four-legged intersections. Their safety is due to the lower number of conflict points, speed restriction for all traffic flows and the difference in entry and impact angles [9].
• Multi-lane roundabouts. These are less safe than single-lane roundabouts, but still safer than other types of intersection [22]. Multi-lane roundabouts are less safe because they have a larger number of conflict points, they are more complex, driving speed is higher, and they have a less sharp entry angle [23].
• Turbo roundabouts (with or without raised lane separators). The road safety effect of turbo roundabouts is almost identical to that of single-lane roundabouts, provided the data are corrected for traffic volume [9]. Converting conventional intersections to single-lane roundabouts (68,1% - 78,7% crash rate reduction) and to turbo roundabouts (76,1% crash rate reduction) generates almost equal road safety benefits [23] [24].

In the urban area, an average 60% of cyclist fatalities occur at intersections, and outside the urban area an average of 49% (in 2015-2019). For cyclists, conventional intersections are generally less safe than roundabouts (also see the section How safe are roundabouts to cyclists?). Intersections with separate bicycle facilities are safer than intersections without these facilities (for an overview of bicycle facilities at intersections, see the section How to make intersections safer for cyclists?)

A second factor contributing to cyclist safety at intersections is the blind spot of (particularly) trucks. Most blind spot crashes are ‘classical’: a truck turning right and a cyclist going straight ahead. Most blind spot crashes occur in the urban area, and especially when the cyclist was riding on a bicycle track [25]. Most of these crashes occur at bicycle tracks at a distance of less than one metre from the carriageway. Half of the blind spot crashes occur at signalised intersections.

Several intersection facilities may enhance cycling safety. These are the most important ones:

• Median island. Median islands make crossing safer for cyclists, because they can cross in stages. This is most relevant at wide and busy distributor roads in the urban area [26].
• Bike box. Bike boxes at traffic lights allow cyclists to position themselves ahead of other traffic, so that they can turn left more safely and easily. They are more visible to motorised traffic since there are no blind spots. That is why bike boxes can prevent blind spot crashes [25] [27].
• Deflecting the bicycle track inwards or outwards. Just before or after an intersection, the bicycle track deflects from or towards the carriageway [25]. This enhances cyclists’ visibility to other traffic. For cyclists on urban distributor roads, the probability of crash on an intersection with an access road is lowest on roads with a one-way bicycle track that is at a distance of two to five meters to the carriageway [26] [27]. A larger distance actually results in a crash increase, although this increase is smaller than in the case of bicycle lanes. For intersections between two distributor roads, the effects of inwards and outwards deflection of the bicycle track and the optimal distance between intersection and bicycle track are unknown.
• Separate green phase for cyclists. Traffic lights at intersections with a separate green phase for cyclists  enhance road safety [18]. This implies that green lights for cyclists coincide with red lights for motor vehicles. Thus, no conflicts between these transport modes would occur if nobody ran a red light.

Which of the facilities above is chosen depends on the volume of passing traffic and on the presence of bicycle facilities on the connecting road sections.

For cyclists, roundabouts are generally much safer than other types of intersection. In the Netherlands, from the 1980s onwards, a lot of intersections were converted to roundabouts, with or without separate bicycle facilities. The effects on (light) moped rider and cyclist safety were also researched at that time. The research showed that conversion of an ordinary intersection to a roundabout resulted in a 60% decrease of casualties among (light) moped riders and cyclists, after correction for the general downward trend in casualty numbers [12][i]. More recent research in Belgium and Denmark shows that, in those countries, the conversion from intersections to roundabouts is less favourable to the number of (serious) injury crashes among cyclists [28] [29]. These studies are, however, hard to compare to the Dutch studies. For example, the number of cyclists, in Belgium in particular, is lower than in the Netherlands, and speed at Danish roundabouts is less restricted by road design than it is at Dutch roundabouts. Thus, it is unlikely that Belgian and Danish results also apply to the Dutch situation.

In the Netherlands, after the conversion of intersections to roundabouts, increases in cyclist safety differ between roundabouts. The size of the crash reduction depends on the bicycle facility chosen. Table 6 shows the effects on the number of (light) moped and bicycle crashes after the conversion. The table compares bicycle facilities in the old situation (intersection) and the new situation (roundabout). The table shows that implementing bicycle tracks at roundabouts decreases the number of (light) moped and bicycle crashes most (89-100% fewer crashes than at the old intersections). In addition, removing bicycle tracks from the old situation (intersection) has the least effect (13-39% fewer crashes at the roundabout) [30]. Moreover, at roundabouts, fewer blind spot crashes occur than at intersections. Only 11% of all blind spot crashes in 2002-2006 occurred at a roundabout [25].

Table 6. Effect of converting an intersection to a roundabout on number of moped and bicycle crashes [30].

Several SWOV reports and a DTV Consultants memo show that the safety benefits of a roundabout are greater when cyclists at bicycle tracks are not given right of way [12] [30] [31] [32]. In 2002, Dijkstra [12] calculated an estimated annual reduction of 52 to 73 inpatients[i], if cyclists were not given right of way at any roundabout with bicycle tracks. An additional factor is that roundabouts are the safest intersection type for cyclists anyway (also see the section How safe are roundabouts to cyclists? ).

Cyclists not having right of way implies that they should yield to motor vehicles on the legs of the roundabouts, cyclists having right of way implies that motor vehicles should yield to cyclists. In general, roundabouts are very safe intersections to cyclists (whether or not having right of way). There are two possible reasons why roundabouts with cyclist priority are less safe [12]. First, drivers could wrongly think they automatically have right of way over cyclists. Second, drivers wanting to enter a roundabout are focused on motorised traffic already on the roundabout, and they have a hard time needing to give right of way to cyclists on a bicycle track as well. Drivers would need to make (too) many observations in a short time, which would facilitate the overlooking of cyclists. Considering cyclist vulnerability in a crash with a motor vehicle (or the great difference in mass between cyclists and motor vehicles), this crash type often has serious consequences, for cyclists in particular.  

Despite the conclusion that, on roundabouts with bicycle tracks, cyclists not having right of way are safer than cyclists who do, CROW recommends to create cyclist priority on these roundabouts in the urban area, but not outside the urban area [36]. This guideline is based on balancing safety, comfort and cyclist traffic flow.

In 2015-2019 an average 26% of fatal pedestrian crashes occurred at intersections. They can be subdivided into crash locations: 36% occurred in the urban area and 7% outside the urban area. The pedestrian share is almost half that of the average share of cyclists in fatal intersection crashes (54%) in that same period. This is also shown by Table 2 at the section How many casualties are due to crashes at intersections?

The degree of pedestrian safety at intersections depends on i.a. the intersection type, presence or absence of traffic lights and zebra crossings.

Intersection type

In general, roundabouts are safer for cyclists than conventional intersections and signalised intersections [37]. This is due to the speed reducing effect of a roundabout, simpler conflicts, and minimum conflict space [37]. As a rule, one-lane roundabouts are safer than multi-lane roundabouts [38]. Furthermore, several international studies show that pedestrians have a relatively higher crash rate, with a more severe outcome, at four-legged than at three-legged intersections [39]. Thus, the pedestrian risk of a fatal crash or serious injury crash is 10% higher at four-legged intersections than at three-legged intersections [40].

Signalised intersections

For pedestrians, signalised intersections are often safer than unsignalised intersections. At signalised intersections, there are fewer conflicts between pedestrians and motorised traffic due to time separation [41]. It should, however, be noted that at signalised intersections motor vehicles may be less inclined to reduce speed during the green phase, which may be more dangerous for pedestrians that start crossing at a later time during the green phase, which implies they partly cross during the red phase. Moreover, motor vehicles running red lights remains a major concern for pedestrians. Conversely, unsignalised intersections force motor vehicles to reduce their speed, since they need to be more careful.

Signalised intersection are, however, safer than intersections with only zebra crossings (see the next subheading) [41]. In addition, pedestrian green phases are sometimes (too) short or coincide with the green phase for turning motor vehicles. This leads to complex situations, particularly when a multi-carriageway road needs to be crossed [42]. In general, turning traffic at intersections (either signalised or unsignalised) is a concern for pedestrians, because they are often overlooked. At unsignalised intersections this is a greater concern, and more so in case of traffic turning left than in case of traffic turning right [42].

Zebra crossing

It is not unequivocal whether intersections with zebra crossings are safer than intersections without zebra crossings [41]. There are studies that claim that pedestrians are given priority by motor vehicles 30% to 40% more often than when zebra crossings are absent [43]. Other studies, on the other hand, do not find this effect [44]. A comparative study by Zegeer et al. [45] for example, found no decrease of pedestrian crashes for unsignalised intersections with zebra crossings. The studies even found an increase in pedestrian crashes when such intersections were passed by more than 12,000 vehicles a day. What needs to be noted is that these are American studies. Dutch zebra crossings may differ from American zebra crossings, e.g. because the latter often use continuous lines as a zebra crossing [42].

For more information about pedestrian safety, see SWOV fact sheet Pedestrians.

Below you will find the list of references that are used in this fact sheet; all sources can be consulted or retrieved. Via Publications you can find more literature on the subject of road safety.

[1]. CROW (2015). Basiskenmerken kruispunten en rotondes. Publicatie 315A. CROW Kenniscentrum voor verkeer, vervoer en infrastructuur, Ede.

[2]. Wegenwiki (2009). Gelijkvloers. Wegenwiki. Accessed on 21-12-2020 at www.wegenwiki.nl/Gelijkvloers.

[3]. Wegenwiki (2015). Kruispunt. Wegenwiki. Accessed on 21-12-2020 at www.wegenwiki.nl/Kruispunt.

[4]. Wikipedia (2013). Bajonetkruispunt. Wikipedia. Accessed on 21-12-2020 at nl.wikipedia.org/wiki/Bajonetkruispunt.

[5]. Wegenwiki (2018). Voorrangsplein. Wegenwiki. Accessed on 21-12-2020 at www.wegenwiki.nl/Voorrangsplein.

[6]. Aarts, L.T. & Dijkstra, A. (2018). DV3 - Achtergronden en uitwerking van de verkeersveiligheidsvisie. De visie Duurzaam Veilig Wegverkeer voor de periode 2018 – 2030 onderbouwd [Sustainable Safety version 3 – Backgrounds and elaboration of the updated road safety vision. Substantiation of the second advanced Sustainable Safety vision for the period 2018-2030]. R-2018-6B [Summary in English]. SWOV, Den Haag.

[7]. CROW-Fietsberaad (2019). Evaluatie discussienotitie Fietsstraten. Fietsberaadpublicatie 32. CROW-Fietsberaad, Utrecht.

[8]. CROW (2019). Handboek wegontwerp 2013: basiscriteria. CROW, Ede.

[9]. Dijkstra, A. (2014). Enkele aspecten van kruispuntveiligheid. Rapportage voor het CROW-project Afwegingskader kruispunten [Several aspects of intersection safety. Report for the CROW project Assessment Framework for Intersections]. R-2014-21A [Summary in English]. SWOV, Den Haag.

[10]. Donkers, E. (2022). Is de fietser nog wel veilig op het kruispunt? Verbeter de fietsveiligheid door het verminderen van conflicten op kruispunten. Paper presented at NVVC, 12 april 2022, Utrecht.

[11]. Churchill, T., Stipdonk, H. & Bijleveld, F. (2010). Effects of roundabouts on road casualties in the Netherlands. R-2010-21. SWOV, Leidschendam.

[12]. Dijkstra, A. (2005). Rotondes met vrijliggende fietspaden ook veilig voor fietsers? Welke voorrangsregeling voor fietsers is veilig op rotondes in de bebouwde kom? [Are roundabouts with separate cycle tracks also safe for cyclists? Which priority rule is safe for cyclists on individual urban roundabouts?]. R-2004-14 [Summary in English]. SWOV, Leidschendam.

[13]. Bout, J. & Olijve, M.J. (2015). Het voorrangsplein: een nieuw kruispunttype?! Onderzoek naar de verkeersveiligheid en de capaciteit op voorrangspleinen. Hogeschool Windesheim, Zwolle.

[14]. Leeden, E.P.J. van der (2012). A comparison between the pleintje, priority intersection & roundabout. Master thesis. Technische Universiteit Delft, Goudappel Coffeng, Delft.

[15]. Reimink, J. & Wiersum, M. (2018). Voorrangspleinen buiten de bebouwde kom. HBO afstudeeronderzoek. Hogeschool Windesheim, Zwolle.

[16]. Janssen, S.T.M.C. (1992). Veiligheid van ongelijkvloerse kruispunten op enkelbaanswegen. Een verslag van een onderzoek voor de Werkgroep 'Ongelijkvloerse kruispunten enkelbaanswegen' van de Stichting Centrum voor Regelgeving en Onderzoek in de Grond-, Water-, en Wegenbouw en de Verkeerstechniek, C.R.O.W.  R-92-35. SWOV, Leidschendam.

[17]. Elvik, R., Høye, A., Vaa, T. & Sørensen, M. (2009). Traffic control. In: The handbook of road safety measures [2nd edition]. Emerald Group Publishing Limited, p. 397-541.

[18]. Ziakopoulos, A., Botteghi, G. & Papadimitriou, E. (2017). Traffic signal installation. European Road Safety Decision Support System, developed by the H2020 project SafetyCube. Accessed on 10-03-2021 at www.roadsafety-dss.eu.

[19]. Janssen, S.T.M.C. (2004). Veiligheid op kruisingen van verkeersaders binnen de bebouwde kom. Vergelijking van ongevallenrisico's [Safety on urban through-road intersections. Comparison of crash rates]. R-2003-36 [Summary in English]. SWOV, Leidschendam.

[20]. Elvik, R., Høye, A., Vaa, T. & Sørensen, M. (2009). Road design and road equipment. In: The handbook of road safety measures [2nd edition]. Emerald Group Publishing Limited, p. 144-333.

[21]. Minnen, J. van (1990). Ongevallen op rotondes. Vergelijkende studie van de onveiligheid op een aantal locaties waar een kruispunt werd vervangen door een 'nieuwe' rotonde. R-90-47. SWOV, Leidschendam.

[22]. Wijnen, W., Weijermars, W.A.M. & Bos, N.M. (2013). Update effectiviteit en kosten van verkeersveiligheidsmaatregelen. Nieuwe schattingen voor elf maatregelen [Update effectiveness and costs of road safety measures. New estimates for eleven measures]. D-2013-7 [Summary in English]. SWOV, Den Haag.

[23]. Fortuijn, L.G.H. (2009). Turbo roundabouts: Design principles and safety performance. In: Compendium of papers DVD 88th Annual Meeting of the Transportation Research Board TRB. Washington D.C.

[24]. Fortuijn, L.G.H. (2013). Turborotonde en turboplein: ontwerp, capaciteit en veiligheid.  Doctoral thesis. Technical University Delft. Delft.

[25]. Schoon, C.C., Doumen, M.J.A. & Bruin, D. de (2008). De toedracht van dodehoekongevallen en maatregelen voor de korte en lange termijn. Een ongevallenanalyse over de jaren 1997-2007, verkeersobservaties en enquêtes onder fietsers en vrachtautochauffeurs [The circumstances of blind spot crashes and short- and long-term measures. A crash analysis over the years 1997-2007, traffic observations, and surveys among cyclists and lorry drivers]. R-2008-11A [Summary in English]. SWOV, Leidschendam.

[26]. Schepers, J.P. & Voorham, J. (2010). Oversteekongevallen met fietsers. Het effect van infrastructuurkenmerken op voorrangskruispunten. Directoraat-Generaal Rijkswaterstaat, Dienst Verkeer en Scheepvaars DVS, Delft.

[27]. Boggelen, O. van, Schepers, P., Kroeze, P. & Voet, M. van der (2011). Samen werken aan een veilige fietsomgeving. Fietsberaadpublicatie 19. Fietsberaad, Utrecht.

[28]. Daniels, S., Brijs, T., Nuyts, E. & Wets, G. (2009). Injury crashes with bicyclists at roundabouts: influence of some location characteristics and the design of cycle facilities. In: Journal of Safety Research, vol. 40, p. 141-148.

[29]. Jensen, S. (2013). Safety effects of converting intersections to roundabouts. In: Transportation Research Record, nr. 2389, p. 22-29.

[30]. Minnen, J. van (1995). Rotondes en voorrangsregelingen [Roundabouts and the priority rule]. R-95-58 [Summary in English]. SWOV, Leidschendam.

[31]. Minnen, J. van (1998). Rotondes en voorrangsregelingen II. Uniformering voorrangsregeling op oudere pleinen, veiligheid fietsvoorzieningen en tweestrooks rotondes [Roundabouts and right-of-way regulations II. Standardising right-of-way regulations on older roundabouts, the safety of cycling facilities, and two-lane roundabouts]. R-98-12 [Summary in English]. SWOV, Leidschendam.

[32]. DTV Consultants (2019). Verkenning verkeersveiligheid op rotondes in Nederland. Memo. DTV Consultants, Breda.

[33]. Soteropoulos, A. & Stadlbauer, S. (2017). Convert junction to roundabout. European Road Safety Decision Support System, developed by the H2020 project SafetyCube. Accessed on 21-12-2020 at www.roadsafety-dss.eu.

[34]. AD (2022). ‘Heilige graal van verkeersveiligheid’ blijkt gevaarlijker voor fietsers dan gedacht. Algemeen Dagblad. Geraadpleegd 03-06-2022 op https://www.ad.nl/auto/heilige-graal-van-verkeersveiligheid-blijkt-gevaarlijker-voor-fietsers-dan-gedacht~a116748a/.

[35]. Bentem, L. van (2022). The impact of infrastructure design on cycling safety. Master thesis TU Delft Civil Engineering and Geosciences. Delft University of Technology.  

[36]. CROW (2016). Ontwerpwijzer fietsverkeer. Publicatie 351. CROW Kenniscentrum voor verkeer, vervoer en infrastructuur, Ede.

[37]. Granà, A. (2011). An overview of safety effects on pedestrians at modern roundabouts. In: Brebbia, C.A. & Beriatos, E. (red.), Sustainable Development and Planning V. WIT Transactions on Ecology and the Environment, Southampton, p. 261-272.

[38]. Brude, U. (2000). What roundabout design provided the highest possible safety? Swedish National Road and Transport Research Institute (VTI).

[39]. Soteropoulos, A. & Stadlbauer, S. (2016). Risk of different junction types. European Road Safety Decision Support System, developed by the H2020 project SafetyCube. Accessed on 10-03-2021 at www.roadsafety-dss.eu.

[40]. iRAP (2013). Road attribute risk factors. Intersection type. International Road Assessment Programme (iRAP).

[41]. Chen, L., Chen, C. & Ewing, R. (2012). The relative effectiveness of pedestrian safety countermeasures at urban intersections — Lessons from a New York City experience. Contribution to the Transportation Research Board 91st Annual Meeting, Washington, D.C.

[42]. Federal Highway Administration (2009). Pedestrian safety at intersections. FHWA-SA-10-005. Department of Transportation DOT, Federal Highway Administration FHWA, Washington, D.C.

[43]. Nitzburg, M. & Knoblauch, R.L. (2001). An evaluation of high-visibility crosswalk treatments - Clearwater, Florida. FHWA-RD-00-105. Federal Highway Administration, McLean.

[44]. Fitzpatrick, K., Turner, S., Brewer, M., Carlson, P., et al. (2006). Improving pedestrian safety at unsignalized crossings. Transportation Research Board, Washington, D.C.

[45]. Zegeer, C.V., Stewart, J.R., Huang, H.H., Lagerwey, P.A., et al. (2005). Safety effects of marked versus unmarked crosswalks at uncontrolled locations: Final report and recommended guidelines. FHWA–HRT–04–100. Federal Highway Administration’s (FHWA), Washington.