NZ Cyclist/Dooring Map Released by IPRU

The following is reposted with permission from the University of Otago Injury Prevention Research Unit. Original article located here.

Vintage NZ MoT Dooring PosterOne of the more common, yet insidious hazards that utilitarian cyclists regularly face in urban areas is avoiding, running into or being struck by a car door (usually on the driver’s side). In some jurisdictions, “dooring” has been called the most common type of vehicle/cycle collision (Munro), although these events are undoubtedly under-reported and the risk will vary from place to place. The NZ Road Code, like many other countries, warns drivers specifically about causing a hazard to other road users from door openings.

Purpose: The purpose of this demonstration project was to determine the feasibility of developing and displaying a publicly accessible interactive web-based map of police reported dooring-related bicycle injuries among New Zealand (NZ) cyclists. This work was performed by the University of Otago’s Injury Prevention Research Unit (IPRU) utilizing data extracted from the NZ Traffic Crash Reports (TCR) recorded by NZ Police and coded by the NZ Ministry of Transport.

From CAS Vehicle Movement Coding SheetMethods: Vehicle movement data coding was used to select which crashes involving a cyclist injury occurring between 2007 and 2011 were the result of collision with an opening door. TCR records include the spatial coordinates of the crash site and relevant movements of each involved vehicle. The coordinates of each crash site were converted into a format suitable for viewing and sharing in Google Maps™.

Results: Between 2007 and 2011, 245 cycle dooring injuries were reported in New Zealand (mean = 49/year). These represented 6% of all cyclist injuries involving motor vehicles (compared to 19.4% in Victoria, AU). Doorings comprised a much higher proportion of adult (age ≥19) cyclist injuries, 7%, versus 2% for ages <19. Two-thirds of the cases were male and most victims were adults (see table, below). The mean age for females was 31.4 years and for males 39.1. About 20% of these cases were seriously injured; two deaths were reported (counted within 30 days of crash, by definition).

Cycle vs. Vehicle Door Collisions, NZ, 2007-2011

Click for Interactive Map of Dooring Incidents

Click for Interactive Map of Reported NZ Dooring Incidents, 2007-2011

  • Mior InjuryYellow pins are minor injuries (such as sprains and bruises).
  • SeriousRed pins are serious injuries (defined by the Ministry of Transport as “Fractures, concussion, internal injuries, crushings, severe cuts and lacerations, severe general shock necessitating medical treatment, and any other injury involving removal to and detention in hospital“).
  • FatalityBlack pins represent fatalities.

Approximate dooring injury crash locations and some clusters along certain roadways can be seen in many of the larger cities (such as Tamaki Drive in Auckland, Victoria St in Hamilton, and Riccarton Rd in Christchurch).

Utilizing the capabilities of Google Maps Street View™ one is also able to ‘drill down‘ (once in Google Maps™ click on the spot marker, then choose more/street view from the menu) and note street level photographic evidence of parking facilities, bike lane presence or absence, and other road characteristics near to each crash site. It is noted that the street-level picture itself may not always represent the way the road looked at the time and date of each incident due to the relative and infrequent timing of the Google Street View photography. The Street View also does not always properly indicate which side of the street the crash took place.

Reported Bicycle Dooring Crashes by Age Gender, NZ 2007-2011

Age Group






5 15 20

0-14 years

<5 9 12

15-24 years

19 24 43

25+ years

53 117 170


80 165 245

NZ Dooring Yr

Discussion: We identified and mapped 245 cycle car door injury incidents over this 5 year period. This is undoubtedly an undercount as avoidance maneuvers leading to injury may not have been counted and police reports are known to miss more than half of all crashes involving cyclists (Meulners, Langley). It is difficult to say from these data whether the rate of door crashes per number of cyclists decreased over this period or the numbers of cyclists decreased (or some combination, thereof).


  • Injury Prevention – Although numerous community informational campaigns have been implemented to address dooring and laws make it illegal to open doors negligently in NZ, the U.S., and elsewhere, such efforts are bound to have little impact. As Munro states in his 2012 review for the Road Safety Action Group Inner Melbourne: “Most education and communication campaigns where car dooring is a component, such as “share the road” style campaigns are probably ineffective as they often run in isolation of other interventions (such as enforcement) and fail the fundamental criteria of being immediate to the behaviour (i.e. car door opening) and intimate (i.e. a direct, personal communication). Furthermore, there is a wealth of evidence from the road safety literature that campaigns which solely provide information, tips or facts are ineffective.” Therefore, effective efforts to reduce dooring incidents should be focused on engineering and infrastructure counter-measures including proper bike lane design and implementation and offering alternative cycling infrastructure and cyclist choices.
  • Bike Lane Design – The key dimension for a bike lane next to a parking lane is not so much the width of the cycle lane, but its reach (i.e., the total facility width) from the curb-side to the outside of the bike lane. For bike lane design guidelines, NZ, looks mainly to Cycling Aspects of Austroads Guides (2011), which uses a desired reach of 4.2 metres (minimum of 4.0m – maximum of 4.5m, page 29). The bike lane itself should be at least 1.5m. The guidelines state: “It is most important to provide a width that is adequate to accommodate parked vehicles, operating space for cyclists and adequate clearance to accommodate the opened door of parked vehicles.” Exactly what is meant by adequate clearance from doors is not stated, though it is clear that in practice many bike lanes do not.
    Typical bicycle/car parking lanes layout (source: Aust Roads, 2011)

    Typical bicycle/car parking lanes layout (Source: Aust Roads, 2011). Click for larger image.

    These are recommendations/guidelines. NZ road planners are not required to follow them nor necessarily update older facilities. This is part of the reason why, in many jurisdictions, actual bicycle lane dimensions are sometimes compromised from the guidelines.

    Severely compromised functional reach next to a bike lane due to grate, debris and pavement lip.

    Substandard widths of reaches and bike lanes always represents a diminished margin of safety. Invariably, these are worsened in the real world by parking and clearance space lost to a large variety of elements such as: Gutter construction, drainage grates, paving overlays, road debris, street damage, excessive vehicle widths, variable door widths, a range of typical parking (mis)behaviours, parking turn-over rates, as well as cycle and cyclist factors such as bike components, lane positioning and speed.

  • Improved Cycling Infrastructure – A 2012 review, commissioned by the Road Safety Action Group Inner Melbourne (Munro), listed the following infrastructure measures to reduce the exposure of cyclists to car doors:
  1. Repositioning or separating bicycle lanes to the kerbside away from the driver-side of parked cars (i.e. “Copenhagen” and protected lanes; see Australian photos, below),
  2. Removing car parking or extending parking time periods to decrease parking turnover and thus door openings,
  3. Encouraging and guide cyclists to ride away from the dooring zone through the use of painted buffers, shared lane markings and traffic calming that encourages lane sharing, and
  4. Ensuring the road space is most effectively utilised by minimizing parking bay width in order to encourage parking as close to the kerb as possible.

Kerbside bicycle lanes in Melbourne

  • credit: NACTO Urban Bikeway Design GuideBuffered Lanes – Buffered lanes are not encouraged in NZ. Older NZ guidelines say that “Cycle lanes next to parking should not use a “buffer strip” …to separate cyclists from parked cars. Any extra width should be provided in the cycle lane.” While research is lacking, that makes some sense. But other places use them with the painted buffers placed either next to parked cars or between the cycle lane and traffic. Buffers might reduce but don’t eliminate dooring risk, though they offer cyclists some flexibility to ride farther from parked cars. Buffered bike lanes do little to prevent double parking which grows as a problem if bike lanes are too wide.
  • Angled Parking – Where speeds and traffic volumes are low and streets wide, angled parking lanes can almost eliminate dooring risks, but angled parking needs to be carefully designed to avoid backing out conflicts.
  • Protected Lanes – Real world bike lane implementation challenges, combined with safety and behavioral considerations, are leading more and more transport/cycling safety and engineering professionals (Furth) and cycle advocates in NZ, AU, US and the UK to call for more use of protected (also known as separated/segregated) bicycle tracks in urban areas.

credit: NACTO Urban Bikeway Design GuideThere credit: NACTO Urban Bikeway Design Guideare many design possibilities (see NACTO Urban Bikeway Design Guide) for protected lanes, with and without parking. The protection can be parked cars, light or heavy duty bollards, planter boxes, or various types of raised infrastructure. If properly implemented (where roadway widths, traffic conditions, cross traffic and parking policies are amenable) these designs drastically reduce door hazards while appealing to a greater number of more traffic adverse potential cyclists.

Recent research (Teschke et al in credit: NACTO Urban Bikeway Design GuideToronto and Vancouver (reviewed here) and Lusk et al in Montreal) suggest that properly designed and implemented separated lanes can be at least as safe as regular bike lanes, while being much more popular to both riders and potential riders in preference surveys (Bohle, 2000, Emond et al., 2009, Jensen, 2007, Rose and Marfurt, 2007, Shafizadeh and Niemeier, 1997 and Winters and Teschke, 2010). Removal of parking is indicated where there are driveways and well before the protected bike lane approaches intersections to enhance sight lines.

Conclusions: Using Google Maps™ to map and display locations of dooring incidents can be a useful tool for looking at specific crash patterns with reasonable spatial precision and added capabilities. Streets with higher dooring risks can be readily identified in several specific urban settings with New Zealand’s small population and five years of data. To our knowledge, this is the first national interactive dooring map, while some cities are using similar tools.

Displaying incidents in this format can help with injury prevention efforts by highlighting the issue and potentially influencing planners and road safety officials in ameliorating problem areas through infrastructure changes while showing cyclists and safety advocates where the local dooring hazard areas are. Communities should aim for a connected cycle network with a variety of safe and inviting choices free of dooring threats for all types of cyclists for major destinations.

credit: NACTO Urban Bikeway Design GuideComments welcome below.

(Map programming compiled by Brandon de Graaf, concept and writeup by Hank Weiss. University of Otago, School of Medicine, Department of Preventive and Social Medicine, Injury Prevention Research Unit). We wish to thank the several reviewers who offered useful comments on this report.


Munro C. Bicycle Rider Collisions with Car Doors; Crash statistics and literature review. Melbourne, AU: Road Safety Action Group Inner Melbourne;2012.
Meuleners, L. B., Lee, A. H. and Haworth, C. (2007) “Road environment, crash type and hospitalisation of bicycles and motorcyclists presented to emergency departments in Western Australia”, Accident Analysis and Prevention (39) pp1222-1225.
Langley, J. D., Dow, N., Stephenson, S. and Kypri, K. “Missing cyclists”, Injury Prevention (9) pp. 376-379; 2003.
Furth P. Bicycling Infrastructure for Mass Cycling: A Transatlantic Comparison In: Pucher J, Buehler R, eds. City Cycling. Cambridge, MA: MIT Press eBooks; 2012.
This entry was posted in Infrastructure, Safety. Bookmark the permalink.

1 Response to NZ Cyclist/Dooring Map Released by IPRU

  1. Rob Ashcroft says:

    I like the design example of Albert St, East Melbourne – but was wondering whether there are examples of similar (pavement/cycle/parking or car lane) in slightly more sub-urban areas where there are more driveways, and whether these have shown to still work well where a person exiting a driveway in their car would first stop prior to cycle lane, then move out an pause at roadside (or something like that)?

What do you think? Comment below

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s