A couple of months ago Auckland Transport posted a video of a near miss near Mt Eden station.

It seems fairly easy to conclude from the video that the primary responsibility for this near fatal accident lies with the pedestrian. AT emphasised this in the accompanying press release:

The near miss happened when the woman stepped out at a crossing despite the warning bells and lights operating…

“We’ve checked the footage of this morning’s incident and we can see the woman checked to her right before crossing but not to her left and that’s where the train was coming from.

AT then goes on to say it’s disturbing that a number of other people crossed before the woman, even though the lights were on and the bells were sounding.

”The lights and bells mean don’t cross. You can’t be using your phone or have earphones on because you might miss the warning lights or bells.”

However, as the NZTA pointed out in their recent road safety ad campaign, we all make mistakes.

And given that we all make mistakes, it is a fairly harsh and unforgiving rail safety design principle that pedestrians are injured or killed if they are distracted, looking at the ground, looking the wrong way or wearing headphones.

This isn’t to say that we shouldn’t be reminding people about the dangers of being distracted.  Following the tragic death of their son, Karamea Matthes and her husband were on the AM show yesterday urging people to “keep an ear out” for trains.  As she says, everyone’s got a device and people are wearing headphones more and more.

Yet the inference from Auckland Transport’s media release is that pedestrian crossings are safe enough as long as everyone follows the rules and they don’t any make mistakes.  Nowhere in the story does Auckland Transport mention that any work is underway to make this particular crossing, or others, safer.

As you can see from the video, if the pedestrian stepped out at the crossing despite the warning bells and lights operating, then this indicates that the warning bells and lights are ineffective at alerting distracted, mistake-prone pedestrians to the oncoming danger of an approaching train.  Some immediate ideas that come to mind are gates that physically close, or led lighting strips on the ground.

In March of this year, Kiwirail sponsored workshops in Auckland and Wellington on making pedestrian level crossings safer.  This apparently led to a trial for vehicle crossings in rural areas, but my subsequent enquiries to Kiwirail and Auckland Transport have drawn a blank as to what is happening with improving the numerous pedestrian crossings in Auckland. If you know anything further about this feel free to add to the comments.

Share this


  1. What about gates that automatically close the pedestrian crossings when the lights and bells start. For safety reasons these would need to be able to be pushed open from the track side to all someone crossing the tracks when the bells started to get out to safety.

      1. If they’re some places they should be everywhere. If they can’t afford to grade separate crossings then automated gates seem like a quick win from a safety perspective.

        1. Don’t think KR can even afford the gates though did add some a George level crossing last year,

    1. +1 to that!

      At KiwiRail, since the current CEO came into office there has been a major emphasis on keeping staff safe at work. After the incident involving Paul Anderson, there has been redoubled efforts to keep non-KR staff safe as well. Also KiwiRail funds (to what extent I can’t recall) TrackSafe.

      Given the above, it seems strange that automated pedestrian gates (Kingdon St, Newmarket being a good example) aren’t installed in more places. If I were to speculate as to why, I’d suggest that with somewhat constrained government funding and the governments general disregard for public transport, KR is not able to fund this directly and probably holds the opinion that AT should fund the gates.

  2. We have automatic closing gates at Baldwin Ave. It makes absolutely no sense to me that we don’t have them in all the stations.

    1. Basically because they’re not cheap, remembering that they have to be incorporated into the track signaling system as well.

    1. I hope they’re going to be better than the ones being trialed at collins ave, in Wellington. Those ground leds are designed for road vehicle usage and are good when viewing from a distance, but not very good when you’re stood above them, which the Pedestrians are.

  3. Seen so many near misses at Avondale station (the crossing closest to St Jude Street). In peak times there are often two trains passing through at the same time. Other times people just can’t be bothered waiting one minute when the train is dwelling on the westbound platform.

    It will only be a matter of time until there is a fatality.

  4. Are there any ungated vehicle crossings on the Auckland metro rail network? (Papakura to Pukekohe excepted)

    1. From what I can tell, within the Auckland Metro rail area, not including past Papakura, all road level crossings have signalised gates for drivers to prevent unintended crossings by motor vehicles. Why is the same level of protection not afforded to pedestrians?

  5. I have a suspicion that the more “difficult” these pedestrian chicanes become, the more people will be tempted to bypass them and cross on the road.

    1. i agree, some people think they are above rules and/or that they will be gone before the train arrives. You can lead a horse to water but you cannot make it drink.

  6. I can’t understand why we don’t have LED lights on the road at traffic lights. I wonder how many crashes that would save. Surely it wouldn’t cost much?

  7. Also what about the option of a fixed camera at every level crossing, and the driver of the train has a live feed of the next level crossing. Probably wouldn’t stop most collisions, but would help with the ones where someone gets stuck on the track.

  8. Railway used to pride itself on its ‘fail-safe’ philosophy. Yet here a minor mistake is a death sentence.

    The railway has shown it can be fail-safe for pedestrians: over-bridges, electric gates, etc, though we hear there’s not enough money for this. What are the priorities? Has anyone actually asked for the money?

    The railway is relying on our most vulnerable to ‘not make a mistake’ and blame them when they do: Our children, our grandparents, and those with all types of varying abilities (many who don’t have the option to use a car). Everyone makes mistakes: You do; working parents; people under day to day stress; or your friend running late for work who is distracted for a moment. Should this be a death sentence in a so called ‘fail safe’ industry?

    The approach needs to be shifted to asking and campaigning for the funding to make crossings safer. Relying on the pubic to ‘not make a mistake’ is a fool’s errand.

    1. ATAP had 600m odd for grade seperations, TFUG plans to get rid of nearly all the Southern ones.

      But Pre-ATAP was very little interest from Central Govt

  9. All of the above is true but there is also the issue of personal responsibility. Half those people didn’t look distracted – they looked like they made a value judgement that running the risk of getting killed was worth not having to wait.

    So you can put in gates etc (and I agree that they should be prioritized) but guarantee some people will still jump the gate

    1. > “guarantee some people will still jump the gate” is a very different problem which shouldn’t stop us protecting all of the people who don’t jump gates. (Some people drive recklessly so let’s not put airbags or seat belts in cars.)

      It is one thing to ‘take responsibility’ if you dent someone’s car or if you hurt someone else, but how do you ‘take responsibility’ for a minor error that happened near a hazard created by someone else when you’re dead?

    2. Personal responsibility? There are gates for all vehicle rail crossings within Akl metro rail area, to prevent unintended crossings by distracted drivers so why not the same for pedestrians?

      1. Personal responsibility in that (as Kevin says and I agree), all but the last woman looked like they were aware of the bells and thought the train was far enough away to just ignore the warning and go across.

        That’s their decision.

        Regardless, I think we should have closeable barriers for peds, but some will jump them. Just as some people choose to drive around vehicle barriers… You can’t argue that people shouldn’t take personal responsibility about their surroundings as they drive, cycle, walk, run etc to keep themselves safe.

        1. I disagree with this hyper-rational and individualistic take on the issue of safety, which ignores the wider social trauma that is caused when people suffer serious injuries.

          Case in point: The drivers of the trains which hit these people offer suffer psychological trauma, which requires ongoing counselling. Similarly for people who might witness someone getting hit: Their emotional state is likely to be negatively impacted. As are the energency services.

          I don’t accept that safety is simply a matter of saying “personal responsibility” and pretending that will work out best for society as a whole, including but not limited to people who do actually make responsible decisions.

          I happen to be a relatively “safety conscious” person and I still don’t want to live in a world where I see people squashed like bugs when they happen to make a bad decision. That could be my child, or my wife that you’re talking about and I personally am very happy for “big brother” to intervene and engineer the world so that that doesn’t happen to them.

          And that’s before we start talking about delays to other passengers caused by such events.

  10. Automatic gates seem like a good idea. Is anyone actually studying it, though? Is there any money going into how to design for pedestrians? Or are we just assuming?

    What are the implications for light rail?

    1. Some studies for light rail here:



      It will be really interesting to see how this is accommodated on Dominion Road. I would think it is going to require signalized pedestrian crossings in a lot of places (every 200-400m?) with some decent phasing or pedestrians are going to be hit a lot.

      1. I had read somewhere that modern LR can brake in the same distance as a bus at 50kmh, in which case it isn’t such a big issue. However, if this is not the case it is a real threat to the proposed LR network, as even with good crossings every 200 – 400m people will still cross the road where they are, especially if the phasing is in favour of the very LR vehicle they are trying to catch.

        1. I doubt that narrow steel wheels on steel rails can stop in the same distance from the same speed as bus wide rubber wheels on road surface.

        2. My understanding is that it is achieved using magnets that clamp down on the rails. Would be good to get this confirmed or denied by someone who knows more about this than me though!

        3. Yep almost all trams have electromagnetic emergency breaks that don’t rely on friction to work.

          A tram can stop much faster than the people inside can handle, like 50 to zero in about ten metres. Hit the emergency stop with earnest and you’ll break bones onboard.

          I don’t know about buses, but my guess is the tyres would lose traction and skid under that kind of pressure.

        4. This brings back memories of high school physics. Something about coefficient of friction between the wheel and road surface. Once the tyres lock up, a car will stop as fast as a truck as mass cancels out of the equation IIRC. Steel on steel would have a different coefficient I think.
          Anyone with more up-to-date physics knowledge care to intervene? This is from the late 90s for me!

      2. Thanks for those links, Sailor Boy. They were interesting, and suggest to me that pedestrian safety is an emerging field for LR rather than one where best practice has already been established. This one was interesting too:


        Certainly one of the key elements seems to be channelling pedestrians as little as possible, but in a way that makes them have a line of sight towards the oncoming vehicle. I can’t imagine this on Queen St. Do you envision fences down the street either side of the LR?

        I’m sure there are people who work in this area at AT, and who will want as much budget as possible to put into good safety measures for Auckland’s coming light rail. I would love one of them to write a detailed post sometime. It seems to me someone will make a decision trading off safety and cost. The more transparent that decision is, the better.

        As with most things, it’s cheaper to build it right first time.

        1. Thanks again. To me, that looks preferable to what we have now, and at 30 km/hr seems fine. Have you tried using Googe Translate? It’s getting better all the time, and I have used it successfully for research papers. At least enough to get the gist.

  11. Cars can get stuck, gearbox might jam, all number of mechanical issues why we’d want a physical barrier at a train crossing. Pedestrians, unless with a physical disability, don’t have these problems. In my brutally honest opinion if you are crossing rail tracks, there is NO way you don’t know you’re doing so in these cases, and it should be personal responsibility to check the way is clear.

    If you just assume it’s ok and something happens, we have an award named after a certain scientist that you can have.

    Also, this reminds me of Sydney’s Darling Harbour where a drunk guy fell into the water and drowned, as I recall doing something he shouldn’t. Then some suggested the whole place be fenced off. At some point we just need to accept shit happens, you can’t and shouldn’t have to protect against all human stupidity to the detriment of everyone else. Thousands and thousands of people walk around Darling Harbour without falling and drowning.

    1. Yes but these crossings are deadly and there are distracted people and children and animals all using them You can’t design a deadly crossing and then absolve the design of all blame when the inevitable fatality occurs. Instead, design for safety.

      This will become more important in the future as we get more trains going more frequently to more places. The time to implement better safety standards is surely sooner rather than later.

    2. It is a blurry line between human stupidity and an accident waiting to happen. By your argument it would be fine to leave live bare wires lying around as long as there is a sign telling people not to touch them?

  12. The safer you try to make crossings the more you encourage rebellious behavior.

    You could try and gate every crossing but it would merely encourage people to find alternatives to make crossings faster. People will jump over fences, cross at platforms, utlilise a nearby road etc to make the crossing faster.

    It all comes down to how we run our society in New Zealand. We are so socialist we’ve managed to create a society whereby people try to circumnavigate the law rather than taking ownership of the situation and conducting themselves in a manner appropriate to the situation.

    1. “The safer you try to make crossings the more you encourage rebellious behavior.”

      Applying this rationale would suggest that providing car safety features such as air bags, ABS, and traction control causes more reckless driving. Yet the data shows there are far fewer deaths (per population or per vehicle) than ever before.

      Source: 3.3 deaths per 100k vehicles in 1980 vs 0.9 deaths per 100k vehicles in 2015 http://www.transport.govt.nz/research/roadtoll/annualroadtollhistoricalinformation/

      Reckless people will always be reckless. This isn’t about them. This is about protecting normal people doing normal every day things from a hazard created by an industry that cannot pull itself out of the 1950’s.

        1. And the data also shows that volvo drivers don’t die in accidents. Check this report http://www.iihs.org/iihs/sr/statusreport/article/50/1/1

          There are 6 car makers that have had no deaths in their cars. So… if that logic works then they’re all driving recklessly but they’re not dying …. so things must be safer.

          So what’s wrong with making the railway safe? are you worried more people will become more reckless? Who cares if that it saves some lives?

          maybe we need to make cars less safe so people take more care.

        2. The really interesting line in that website (thanks Bill M for the link) is that apparently, Subaru Legacy owners have had zero deaths in crashes (presumably USA?). I have no figures at hand, but from memory, Subaru Legacy is one of the most frequently stolen cars in NZ and therefore also one of the most frequently crashed and set alight, and again, from memory, there has been a fair amount of people (mostly idiot car thieves, hooning around) killing themselves in those Subarus in NZ. So its not just what you’re driving, but the mentality of the person driving.

      1. Your analogy doesn’t hold true as every feature you mention is invisible to the driver. We are talking about highly visible changes designed to change user behavior. They will change behavior all right.

        It’s amazing how one video can influence the opinion of the masses. If I show you a video of a cyclist riding into a parked car would you back a change to cycling to include a compulsory warning device on all bikes to warn riders of dangers ahead?

        1. How many cyclists have died in the last ten years from riding into parked cars? How many have died at rail level crossings?

          And if you think that highly visible changes to change user behavior lead to more deliberate rule breaking, then surely boom gates at road level crossings would be worse than flashing lights? Or speed bumps would actually cause people to speed.

        2. In answer to your question: in the past ten years there have been two of each killed by trains or the back of a parked car…

        3. The car analogy works just fine. Features don’t need to be visible to improve safety. Some of the best safety isn’t obvious.

          Grade separation allows people to do their normal things like being a bit disabled or a bit deaf or a bit blind or in a bit of a hurry…. and they won’t end up flat as a pancake.

          We just need to somehow get all the people of all the ages to have the experience, attention, eyesight, hearing, mobility, patience, and alertness that you all do. Is there a training course we can do?

    2. Too cynical, 99.9% of people aren’t interested in rebellious behaviour. There will always be idiots and neerdowells. Gating crossings looks like a great idea, I would go further and get ground based flashing LEDS and a focused electronic sound jammer that interrupts devices with an annoying schreeching when crossing lights are flashing.

    3. Ha Ha, We’re so socialist we’ve had a tory government for the past 9 years. Go find someone else to blame for your personal angst.

  13. Personal responsibility suggests you look where you’re walking, not in a dream world with a cellphone and headphones. If you’re so stupid you only check one way and ignore lights and alarms how on earth do they cross the road safely?

    1. Roads have pedestrian crossings where a driver stops for pedestrian crossing. You are right… that is what we need on the railway… though cars can stop pretty quick.

      If we don’t cater for people with headphones, does that mean we don’t care about people who are bit deaf? – which includes half of my tradie mates.

      ha ha the funniest bit is you said people should look where they’re walking. If they do that they wont see a train coming at them. Keep looking straight ahead bro. You’ll be sweet.

  14. Irrespective of people’s thoughts on personal responsibility or looking after societies more vulnerable members, one thing is for sure – the increased frequency of trains will necessitate a solution.

    A train each direction every five minutes, will make these crossings so inefficient there will need to be alternatives to make sure people don’t miss their train.

  15. That lady was extremely lucky!
    I 100% agree that you can’t achieve 100% safety, as there is always going to be the hazard of a 30 tonne (or however much a train weighs) lump of steel moving at speed, within a close proximity to people.

    However, I also 100% agree that there is more that could & should be done at pedestrian rail crossings to make them safer. The bells are loud enough, they have flashing lights. I would say gates would be the most logical thing to implement to reduce most if not all deaths. Adding more lights (be in it the pavement) or more noise would still not stop people who are distracted. But a physical barrier would. I’d almost argue that the loud bells could have volume turned down a bit if barriers were installed.

    The question of timing of when a barrier closes would be interesting … As some people are in a hurry and if a barrier was to close 1 minute before the train arrives, then that would encourage people to jump / duck the barrier.

  16. KiwiRail (with NZTA) are actually doing quite a lot about level crossing safety at the moment (I have been assisting them with this work), with a particular emphasis on ped’n/cycle crossings – whereas road crossing crashes have been tracking downwards, path crossing crashes have remained static or getting worse.

    A new level crossing risk assessment process has just been developed, and the first training course for assessors was recently held. The idea is that all future new/upgraded crossing will require a formal safety assessment process that will determine whether the proposed crossing treatment will actually reduce existing risk or not. This will more clearly inform what treatment options are needed to achieve the desired risk level, e.g. automatic gates, chicanes, grade separation, etc. In the case of Auckland, ultimately to achieve a sufficiently low level of risk, virtually every crossing will require grade separation in due course.

    Ped/cycle crossings have previously not had a lot of specific technical advice on their planning and design in NZ; the sudden flurry of interest in Urban Cycleways and NZ Cycle Trails has required more than an ad hoc approach. New guidance now highlights a number of key planning/design principles for crossings, which include (1) Seek AWARENESS (by users) of crossings and trains and (2) Seek COMPLIANCE by users when crossing. These deliberately reflect those Safe System principles that we are used to in road safety, i.e. humans make mistakes and silly decisions. Awareness is particularly a key challenge for distracted users, hence the current interest in tools like in-ground lights and chicane barriers to catch attention.

    Introductory workshops on these new risk & design guidelines were delivered to industry back in March, and subsequent feedback has been taken on board to revise the draft guidance. The new versions will soon be launched for general use (which reminds me that I’d better get on and make the final changes…).

    1. Thank you very much for that information, Glen. Does the assessment process involve monitoring what’s going on by counting pedestrians, the proportion with a mobility aid, cyclists, children, the proportion who don’t look in the correct direction, etc? I would be interested to read the guidelines when they are finished – will they be publicly available?

      1. Counting the numbers and types of users is a key part of the assessment process, and not something that’s historically been done (in the database of all railway crossings, it’s amazing how many currently have exactly 100 peds per day…). For existing sites, on-site observation of behaviour is also expected. Like all industry guidelines, the next version should be publicly available.

  17. If I recall correctly, the young man killed at Metcalfe Rd ran behind a departing city-bound train and into the path of a Swanson train, which he likely wouldn’t have heard because the bells were still ringing.

    As a regular user of the gated crossing at Ranui, the emergency gate is regularly used by those catching city-bound trains, while a Swanson bound train is at the adjacent platform. A Swanson bound train closes the gate as it approaches the platform, and keeps it closed during the train’s lengthy dwell time, and doesn’t open until several seconds after the train has cleared the crossing. The city bound train will then close the gates before it rounds the bend some 400m away, potentially meaning the gates can be closed for up to 2 minutes at a time.

    Have also seen people cross the tracks at Henderson and jump over the Railside Ave fence.

    1. Fine idea, but have you got a billion bucks for that? Or do you have a particular budget in mind to slash to do it.

      I’m mean, in a perfect world we’d just drop the RoNS budget for it, but we don’t live in the perfect world just now.

      1. That’s only if you’re grade separating roads as well. Let’s just fix pedestrian/cycling access. Motorists can wait or find a grade separated route. Save a small fortune.

      2. Yes, the east west link, at 1.8 billion, more than enough and in the meantime we can have a decent conversation about how to solve issues around onehunga.

  18. all these extra safety measures come with a cost which at end of the day gets passed on to the users, which could have the effect of making trains less competitive with other more dangerous modes.

  19. This is all very well. Make the railway as safe as possible etc etc. But what about the roads? How many more casualties occur with pedestrians crossing roads? Or worse, with errant vehicles driving onto the footpath? Why is this not receiving vastly more attention than rail-crossing safety?

    In a nutshell, 300+ die from road traffic accidents every year. 10 times more are seriously injured. A small handful die or are injured on the railway every year, and yet only the railway is expected to take meaningful action to be safer.

    The roads carry on as usual. So do the deaths and injuries. We know that 300+ will die this year. 3000+ will be seriously injured. What meaningful changes are we making to stop this? To make sure it won’t happen next year? Answer NOTHING. It will happen next year.

    This glaring disparity in the tolerance of road casualties compared to rail is the crux of what is wrong with transport policy, pretty much world wide in fact. Roads, road-vehicles and road-vehicle-operation is where drastic changes are needed.

  20. “Expect Trains”. Trains can appear at any time, from either direction. This is what we promote at KiwiRail. If we use designated crossings, we ALL still need to understand the potential for underestimating the consequences when a rail crossing is taken for granted.

    I agree with the NZTA, that yes we are human beings, and yes we do make mistakes. But we still do need to ensure that the public is educated about the risks and dangers associated with crossing the tracks. Especially within our schools for our younger people who are more likely to mis-judge their surroundings at a rail crossing.

    The worst, most horriblist, and all to common thing we have happening, is an unauthorised public education taking place from these actual incidents taking place. In example being this awesome, precious, young man with a very bright future running out in front of a train accidentally.

    1. Children will pick up on how to cross a railway by watching how adults do it, just as children learn how to cross roads from adults. An education programme to adults is a very good idea, but it needs to accompany a built environment that allows adults to easily accomplish what their sense of responsibility would have them do. Any education programme in schools – if adults aren’t following the recommendations – will soon be undermined by what adults do, and will add to the load we are already putting on children to fix our own bad habits.

      I don’t have a lot of experience with crossing railways. With roads, most adults don’t model good crossing habits to children. But then, Auckland’s built environment is not conducive to walking, as has been commented by many people on this blog. We are in a city designed around the car.

      From the psychology paper I quote above (about design for LR):

      “The key is to design things intentionally that really channel attention and behavior through designing from a human-centered systems design perspective,”

  21. Lets follow melbourne, when I went there to live, I was surprised to see these amazing safe gates that close for a minute or two when the train rolls by, doesn’t look that expensive to install and surely we can’t put a price on more and more peoples lives

    1. The Level Crossing Removal Authority in Melbourne has a huge budget of billions of dollars and a 30 year plan to close and/or grade separate many crossings. It would be possible to do the same in Auckland but there is no funding available for this. It may also be possible to install gates at many pedestrian crossings but this is not tgat simple as they need to be connected to and work in conjunction with the signalling system. Unfortunatly this isn’t a cheap option and there is currently no funding available. This makes the trial of easily installed LED warning lights into the approach a real positive as its potentially a cost effective and simple addition which will make a real difference.

  22. UPDATE: I’ve been forwarded the latest newsletter from Tracksafe, that contained this:

    Pedestrian LED lights have been installed at Collins Avenue in Tawa and are aimed at alerting pedestrians who are looking down at their phone or device. The LEDs activate whenever an approaching train is detected.. Other measures incorporating LED technology are being developed and planned for trialling in 2017/18 at four locations in Wellington and Palmerston North.

    There’s also this article from last year:


    A Tawa railway crossing is about to become one of the safest in the country, with KiwiRail announcing that it will be the first site to have voice message units installed.

    The system, to be installed at the Collins Avenue crossing, aims to reduce the likelihood of a pedestrian walking out in front of a second train just after a first one has passed.

    A voice message will be activated to warn them that another train is approaching…

    …The crossing had been chosen as the first installation place for the technology, because it already had a second train detection system.

    Currently, two hooters sound and a visual display unit advises pedestrians that a second train is approaching.

    Once the system had been up and running for a while it would be evaluated and hopefully installed across the country, Cook said.

  23. The kids still kick the barriers up and keep going even with flashing lights and bells have seen this on numerous service maybe the assume every train stops and do t realise there are expressed but as the emergency brakes take approximately 30 seconds after being applied it can be to late they run in front of 2 trains at paraparaumu as well

  24. A cheaper concept would be a spring loaded swinging gate that can be locked magnetically (like the security locks on a building). A person would normally have to pull the gate open to cross.

Leave a Reply

Your email address will not be published. Required fields are marked *