With the large transformative PT infrastructure projects, such as the City Rail Link and Light Rail, many years away, I’ve been thinking a lot recently about the importance of improving buses as a way for Auckland Transport to continue to deliver strong growth on the network. This has focused on improving frequency and reliability. In this post I’m going to look at how particularly fixing the latter can also make our buses faster.

Getting our buses to run faster has two main benefits

  1. It makes buses more attractive and more competitive with other modes. This means more people will use them.
  2. It can enable more efficient, allowing the same number of buses and drivers to make more runs, thereby increasing frequency and capacity. Alternatively, the same timetable may be able be to be run using fewer buses, thereby saving money.

Perhaps the main tool to achieving faster buses is the same one that will also improve reliability, AT’s Connected Communities programme (formerly the Integrated Corridor Programme). Although as I said in the reliability post, I’m concerned at ATs record of delivery on stuff like this. On this programme AT say:

As part of this RPTP, AT is also looking to apply some of the advantages of the RTN to the Frequent Transport Network through the Integrated Corridor Programme. This Programme will seek to extend bus-priority for the full length of key FTN routes, improving average speed and reliability and reducing operating costs.

Auckland has constrained arterial corridors and there will be trade-offs to be made around competing uses including general traffic lanes, cycle lanes, parking and median strips. AT will design and deliver whole-of-route bus priority on the FTN where:

  • current and planned services experience inconsistent travel times due to congestion
  • where travel-time savings and patronage levels justify the cost of delivery
  • where capacity exists, or new services are planned that can leverage priority infrastructure to deliver patronage growth
  • if reallocation of road space is required, where expected patronage gains are sufficient to ensure that bus priority implementation will increase overall people throughput along the corridor.

How fast are our services?

To start with I thought it would be useful to see just how fast our buses were.

If you’ve looked at a timetable in any depth you might also notice that routes tend to run faster off-peak than they do during the peak. This is because off-peak there is less congestion on the roads and there are generally fewer passengers, speeding up dwell times.

Theoretically, by providing that whole-of-route priority it should mean that buses at peak times can get closer to at least matching off-peak times. So I thought I would compare the two.

Using AT’s transit feed information I calculated how long each trip took and the distance of the route. There are a lot of services so for the purposes of this post I’ve only compared the ones that run regularly all day, the frequent and connector services. As route lengths and times can vary significantly, to better compare them I then looked at how the average speed of these services in the morning peak and compared that to the how long they take in the middle of the day.

The results quickly showed some notable trends, particularly related to services that go to/near the city centre so in the chart below I’ve split them out from the other services. I’ve also included the trains for comparison.

Some of the things that stand out to me are:

  • The slowest (scheduled) bus route is the City Link, traveling at a zippy 8km/h – this will be unsurprising to many who will have found they can walk up Queen St faster than those buses.
  • Most of the green routes are very slow with the exception of a few faster ones, which are generally services from the Shore or Northwest that use the motorways for a portion of their journey. It is particularity notable that many of the isthmus corridors are amongst the slowest, averaging just 13.5km/h at peak times. These tend to be the routes that have some existing bus priority so it is scary to think how much slower they would be without it. The fact that city routes are slower would be further reinforced if I were to include peak only shorter running services, like the 22A, 24W or 27T as the average speed on all of these routes (and other similar ones) was slower than the longer all-day routes.
  • For those routes that don’t go to the city centre, average speeds at peak times are nearly 50% faster at just under 20km/h.
  • Off-peak speeds are around 3.6km/h faster. This may not seem like much but as the next chart which just looks at the difference in peak vs off-peak speeds shows, this represents an average speed increase of about 18%. For many routes, this means a travel time saving of 10 minutes or more. That kind of saving could be significant and in some cases, may be enough to allow frequencies to increase without costing AT any extra.

On top of just giving buses whole-of-route priority, there are also a couple of other things we should consider doing to speed up buses.

Improving dwell times

A couple of tools that we should look to adopt have helped speed up buses that have been successful in other cities

Off-board fare payments.

Now over 90% of all bus trips are made using HOP. On these frequent corridors, for anyone needing a paper ticket we should require one is purchased prior to the bus arrives which will help speed things up. For the Northern Busway we should aim to go a step further and like we would a train station, with people tagging on at the platform instead of on the bus.

Allow all door boarding

Allowing people to board buses using all doors can have a significant impact. For example similar measures introduced on some San Francisco buses saw a 2% improvement in travel times. This kind of change would be helped by the fact that over 90% of buses users now do so using a HOP card so they can easily tag on to the reader at the rear door.

Wider Stop Spacing

One of the things that stood out to me in the graphic I created recently to show the walking catchment of light rail on Dominion Rd was just how close some bus stops are. In many places, bus stops are barely a few hundred metres apart and that frequent stopping can be a big source of slowness.

Fewer stops means buses can should be able to travel at higher speeds for longer thereby be faster. The trade-off is that some people will have to walk further to get to a bus stop. An example of this is shown below for Sandringham Rd where in the 3.8km between Stoddard Rd and Kingsland Station there are currently 14 stops, giving an average distance between them of just 271m. If we instead reduced the number of stops to 9 it would see the average distance increase to about 420m. The red represents a 400m walking distance to the current bus stops while the blue a 400m walk from some potential stops.

There are definitely some areas that won’t have as much coverage based on these station locations but how much faster does it make buses?

Using the speed information above I’ve pulled together this chart to compare the average speed of buses and the average distance between stops. I’ve the same breakdowns from above for consistency and plotted them as separate series. What this shows is that there is some benefit from having longer stop spacing, potentially by up to 1-2km/h overall. Although this may be included in the peak/off-peak comparison as buses won’t be stopping as much to pick up passengers.

Either way, fewer stops served by more frequent and faster services seems like a good trade-off to me. Stop spacing is definitely something that the Connected Communities work should be looking at although I’d note that anyone who thinks it is easy to move a bus stop has not tried moving a bus stop.

Share this


  1. Wow amazing work. I hope that Auckland Transport does this kind of analysis – we need them to be forensically going through these slow routes and making important tweaks to speed the buses up.

    1. AT are not that clever. Their recent installation of several traffic islands to protect cycle lanes in Papatoetoe town centre is proof of that.

      Buses are now stuck in traffic jams for significant periods during peak time while the cycle lanes remain empty.

      Here I was thinking buses would be prioritized over non-existant cyclists.

      1. Yes they have taken out the slip lanes on the St George railway over bridge however they had to because of the double fatality last year when the 380 bus killed two pedestrians. I agree there are very few cyclists on the cycle lane but there are some and less cyclists on the footpaths. Anyway its safer now for everyone which after all is the most important bit. Between 5.00pm and 6.00 pm there is congestion on Station, Wylie and Shirley roads which will hold up buses. However much of this traffic is trying to avoid using state highway 20 and may go another way if congestion worsens.

        1. I’ve heard from a regular user of those buses that they are up to 15 minutes behind schedule during peak times.

          The 380 bus accident was on the Wyllie Rd / Puhinui Rd intersection and involved the driver running a red light.

        2. Your right it was Wylie / Puhinui road intersection.
          I am happy with the lane changes made on the St George Railway overbridge as it slows traffic down and makes it safer for pedestrians. The bridge always has being a bottle neck though and it is holding up buses during peak traffic periods. I suppose we will just have to live with it. Its being the same for the last 20 years I have lived here. Not to sure the cycle lanes are making things worse. A few years ago a women was knocked down by a car while trying to cross station road. She was okay but it has made me weary of the road and anything which slows traffic down is a good thing in my opinion.

  2. Excellent post. There are some very low hanging fruit that Auckland Transport could pick and that would make massive improvements and without costing significant sums of money. The 18 service has been touted as a frequent, reliable service on a “turn up and go” basis and while it’s OK heading into the CBD, leaving the CBD is excruciatingly slow, if the bus actually turns up on time. Speeding things up on Great North Road would help – although AT has not progressed matters at all with bus lanes on Great North Road, it could remove some of the bus stops to speed up journey times. It would be great if AT actually had the guts to improve bus reliability in the CBD but I don’t hold out much hope.

  3. Matt, this analysis is so interesting. Thanks.

    I haven’t seen anything from the Connected Communities team yet. I’ve no idea how they’re working, but they could start work now, using trials. The information they could compile from choosing a couple of corridors and altering both the stop spacing and giving some bus priority – just for whatever length of time doesn’t need to be consulted on – could then really inform the whole programme.

    In any case, I hope they will start a PR programme showing us what they are doing.

  4. No one’s seen anything from the CC team yet and I doubt we will for some time yet. One of the issues with it is AT have decided to hire 12 different consultancies to work on 15 different corridors. So that’s 12 different approaches and interpretations of business cases and design approaches etc. The whole thing is set up to fail

    1. WTF!? Designed to fail? You’re saying that again. Who is responsible for this????

      And does the board know? I mean, good governance would have questioned such a stupid move.

      Oooh I am hopping mad.

    2. While it would appear to be more wasteful maybe the wealth of creativity will help to identify more solutions?

      Big problem seems to be getting them past the car committees

      1. The ideas and creativity have never been in short supply. Am I wrong that this could:

        – waste money on work repeated by many different people,
        – create unhealthy competition between designers each wanting their own pet designs to become standard,
        – create inefficiencies by not having standardised designs to use in PR, tender documents, construction, maintenance contracts, enforcement guidelines,
        – weaken the arguments for the best designs because only some consultancies thought they were necessary,
        – delay, delay, delay?

        I’m wondering if it’s some sort of attempt to remove legal responsibility from AT – “all these consultancies say X,Y,Z, so we must be justified in deciding to do so”??

        1. Nah you are not wrong. The overriding factor is AT senior management and what it does with the work it commissions. How the actions of those power structures can be brought in line with actual community interests is beyond me, I just tweet at them.

  5. It’s been stated many times previously but making it compulsory to give way to buses pulling out from a stop would be huge.
    Is there some reason it has not been done?

    1. Requires central government / NZTA to make changes to the land transport rules. AC and AT can lobby for it to happen but they can’t do anything else.

    2. In Christchurch buses will take you out if you do not give way to them. It may not be legally required, but it works.

    3. So how can we influence those who prepare that legislation to get it done?
      (Make it compulsory to give way to turning bus in front of your vehicle.)

      1. So how can we influence those who prepare that legislation to get it done?
        (Make it compulsory to give way to turning bus in front of your vehicle.)

  6. This post is very well done. However, I wonder if any of the AT senior staff or their consultants actually travel on a bus at peak time?
    The solution to speeding up services are manifestly obvious if anyone takes the time to watch what happens at intersections along bus routes. Other road users take it upon themselves to creep through amber/red lights and block the passage of buses until there is movement at the next phase of lights. Stand down in Halsey St out teh AT offoce no less and see the congestion caused by those turning out of Viaduct Harbour Ave/Gaunt Street – often 4 x light phases needed before bus can make it through the intersection. This is repeated at nearly every intersection with folks creeping across the bus lane to try an enter the traffic lane. Endless delays that compound to a significant impact on the schedule results. Perhaps a simple promotional campaign to remind people of the road code followed by ruthless enforcement will change attitudes? I suspect that is too obvious to those who can’t see the wood for the trees. That fact that Mr Ellison of AT can waffle on about road safety but yet do nothing to change the manifestly dangerous behaviour of so many motorists speaks volumns.

    1. Other thing i would also note that’s prevalent on Custom’s St, is traffic creeping over the green/orange light when there’s no space beyond, and getting stuck in the middle of the intersection, blocking perpendicular traffic coming as the phase changes. This needs better enforcement too.

  7. I still don’t understand why more of the central isthmus bus routes (especially Dominion road) don’t use Ian McKinnan / Upper Queen Street instead of Symonds Street. It would be significantly quicker and it would reduce bus congestion on Symonds Street (without spending $3 billion on light rail). It is wide enough to have bus lanes the entire length, and while I understand that retailers wouldn’t want buses on central Queen Street, Upper Queen Street is hardly anything special. Does anyone know why this hasn’t been considered?

    1. The same reason NZTA won’t fix the intersection at Warkworth. They don’t want to introduce a low cost solution that would remove the justification they have cooked up for a major project.

      1. I suspect you are right. Although this was obvious before AT and before light rail was even considered (unless the light rail plans go back further than we know).

        1. When the Dominion Rd/New North Rd interchange was built no provision for trolley bus wires were made so the buses began using View Rd instead. It’s just stayed that way and no-one probably wants to change it lest they face resistance from View Rd residents.

        2. I guess they need to have it one or the other to keep the 25 service simple, rather than having a mix of routes down Dom Rd. And they chose to connect the frequent route to the rail network at Mt Eden. Good tradeoff IMO, and it’s one of my regular routes

        3. They are making the majority of people spend another 10 minutes on the bus so a small minority don’t need to walk 5 minutes to Mt Eden Station.
          This is the problem with buses – there always seems to be a reason to have them zigzaging around trying to pick up and drop off as many people as possible. And then it takes 40 minutes on the bus or 10 minutes to drive (off peak that is).

    2. There is other catchment to consider at other times, those 252/3 are express buses.
      Also consider, most of the Onehunga sourced routes, City Link, SkyBus. Many Western buses use the upper part of Queen St partly due to CRL works, we now have K’rd Street upgrade going on as well.

    3. Because they’d be bypassing a large chunk of the universities, which is where a lot of the routes’ passengers come from.

  8. My observation as a regular user of buses around the Papatoetoe and Manukau area is that the high number of bus stops is not significantly slowing the buses. Take Lambie Drive and Carruth road for the 380 bus. Its not like the bus has to stop at every one. I wondered why there were so many stops then I realised how much it cut down the walking distance for all those potential passengers who live on side streets. It may be different elsewhere but I think AT got it right when they implemented the new network. Every day I am seeing increases in bus usage. One more thing the more bus stops the less on street parking on the streets of the frequent bus routes.

  9. I’d think the biggest holdups are traffic lights, town centres without bus lanes and motorway interchange areas without bus lanes.
    The 18 strikes all these and there is probably little that can practically be done to improve it without some serious demolition or obstruction to general traffic.
    Waterview is probably the next length of Gt Nth Rd to get bus lanes.

  10. Agree about reducing the number of stops on routes. Shakespeare Road in Takapuna is probably one of the worst examples of excessive stops over short distances (it’s flat too) and means you miss light phases, so the time delay is exacerbated. When I suggested reducing the number of stops to the AT people at the time of the new route planning (which for me personally has been a big backward step), the response was negative because apparently it costs $5000 to eliminate stops, so… much better to keep them. I wonder who dreamt up that number and logic?

  11. But also, improving routes by eliminating circuitous detours will improve speeds too. The 856 going to Takapuna wastes a lot of time backtracking from Smales through the hospital entrance on Shakespeare Rd. Detour you ask? Well it is because I have yet to notice anyone getting on or off the bus at the hospital stop (noting there’s a stop outside the entrance on Shakespeare along the way to Smales anyway).

    1. ACB
      I have a different view. I would much rather have the buses run through the hospital and have people use them. One of the reasons that they don’t is that the DHB heavily subsidises car parking. And as one Board member said last year, why should they? Why should we put precious health resources into parking rather than provision of care? If that member is standing again he will have my vote.

      1. Disagree, buses that terminate at Smales should be extended to the hospital, buses passing through Smales should not detour to the hospital.

        1. Interesting thought, which ones are those do you know off-hand, bit hard to tell actually on the maps as so much passes/terminates through these two points.

      2. Don’t get me wrong I’d rather people use the bus to get to the hospital too (and the DHB is crazy to spend money on carparks) but that’s not a reason to persist with a route which benefits effectively no one at the expense of other users. Particularly when other options are available as the following post suggests.

  12. As a regular user of the Sandringham Rd buses, the problem isn’t the number of bus stops on Sandringham Rd. For the most part the buses move quickly down most of Sandringham Rds, although there are a couple of bottlenecks – one is Edendale School where there is no bus lane, and there are a lot of parents dropping off kids in the morning and so it gets congested. The other is the new speed bump by Ethel St, and the loss of the bus lane around here. In addition, the promised 5 minute frequency just doesn’t happen after about 8:15, so you end up with fuller bus stops and you have to wait longer – if the buses came when promised there would be less time at the stops.
    On the way home the problem is in two main places. First is the entire central city. All the work that is being doing is basically making it difficult to travel across the city which is where all these routes go, and it’s slow to get up to Symonds St. The other is where New North Rd goes under Dominion Rd, it’s a single lane under the bridge and without them doing something major to widen the bridge this will continue to be a problem. Once you clear these two obstacles the run is mostly smooth.

  13. Great piece Matt. One must remember, in the case of Dom road, those bus stops are placed in the approximate vicinity of the tram stops. They are their because because humans find them to be ideal locations to board and alight from public transport.

    1. I’ve searched and failed to find the locations of where the tram stops were. In the archives I found community petitions submitted to the authorities to have stops and shelters added here or there, but I never found a map that actually plotted out the stops along the routes.

    2. ‘Humans find them ideal places to board and alight’ is true of every single metre of isthmus arterial routes. No one is arguing that there arent at least some people who are better of with stop x retained. The discussion is about how much we should prioritise convenient boarding and alighting relative to faster average vehicle speeds.

  14. Good post Matt. I can see real benefits in investing heavily into the buses, both in terms of modernising the fleet so we have lower emission buses, and in improving bus lanes, frequency and facilities. These improvements would completely change the service down Dominion and Sandringham Road and could be delivered relatively quickly (bearing in mind we are dealing with bureaucracy) in comparison to light rail which is a decade away at best. Best of all it could be all done for under $500M (there’s no science behind that but you would have to think that would be pretty realistic) leaving some $4B plus in change to make other investments into PT.

    1. If you made all of those bus improvements and had $4b sat around for PT your best option to spend it would be SW light rail. We should do all of this bus lane improvements but it doesn’t change the fact that the city centre cant handle any more buses.

  15. There’s a lot more to this than just bus-lanes and HOP card swipes. Why is the bus so slow?

    Here’s an example … last week I went into town to meet someone for lunch. I planned to use public transport, leaving from Devonport for a lunch meeting at the museum. My options; take the buses (AT’s route planner says one hour 45 minutes), try the ferry (about an hour but stupidly prohibitive at nearly $30 round trip), and driving a car, which took about 30 minutes and also had the lowest cost (cheap old 1.2L car).

    Why is the bus so slow compared with my own car, when the route must be broadly similar? I understand stopping for passengers will add maybe 15 minutes or so, but an extra hour each way, not even including an initial waiting time to cover infrequent buses at the start? Something is absolutely stuffed up here, and it must be at a basic level like timetables, frequency, network design, something fundamental that other cities seem to be able to get right. This needs fixing up if we are to achieve significant mode shift toward public transport. And we need to do that.

    1. Yes all of the above. Frequency, right of way but also sprawling less dense cities will have less than par services all things being equal.

  16. Given buses are going to be the mainstay of PT in Auckland, and I say that with deep regret as it will forever hold back the potential of PT, something has to be done but to be honest, given bus designs, there is little hope apart from your suggestions and even then it will be minor.

    Ideally, light rail would have been brilliant but you know we can’t ride Phil Twyford’s ghost trams so that’s out, probably for all time, unless someone in the Greens or Labour wake themselves up from their self imposed torpor on this subject and can sort their shit out rather quickly. Realistically though we will have to put up with another bout of Nationals road and motorway solutions before that opportunity ever presents itself again.

    Otherwise a few suggestions;

    1) Bendy buses, quicker to load and unload.

    2) Ban the physically impaired from the upper deck of double-deckers. In the perfect world, we all have as much time as a gold cardholder to get on the bloody bus and into a seat and to realise they have arrived at their stop and actually get off on the same day at the same time as searching for and overcoming the complexities of swiping a hop card, every time, but this ain’t a perfect world!

    3) Electric buses. Quick off the mark far more than anything we have out there at this time.

  17. We should get smart traffic lights that can sense bus.

    If a bus is approaching , the traffic light should turn green.

    Also cars cutting in front of the bus on a bus lane but not turning left should be fined. Bus should have a video camera in front and send those videos to enforcement and issuing bus lane fines.

  18. I remember as a kid back in the 60’s buses departing the city during peak hours had ticket sellers standing on the road selling tickets for the buses making it faster to board . And they also had them on board the buses doing the same .

    But I think alot of them were from the days of the trams and the ARA didn’t want to sack them and it help to get the services moving faster as all you had to do was show the driver the ticket , also you could buy them on board at the same time .

  19. Traffic light priority for buses is worth investigating. Why not start prepaing now for light rail along Dominion Rd by allowing buses to use a quided right of way, without expensive rail lines, just as the trams did and will do in the future. Passenger refuges would have to be provided. Shop owners would be happy and buses would have priority.

    1. Auckland already has a traffic light priority system on it’s buses, it’s the same system that runs the real time tracking and passenger information displays. Unfortunately it’s largely switched off because we run too many buses.

      The problem is we already run far too many buses on routes like dominion road for light priority to work, there is often more than one per light phase. If you tried to give them all priority the direction buses are travelling would never be able to change away from green, and the side streets and turning movements would be constantly red!

      Light rail works much better with light priority because you have one very high capacity vehicle every five minutes which means you can give it full priority and still have another two signal cycles for the other movements to catch up… instead on one regular capacity vehicle every minute or two where you never can.

  20. Another delay is caused by the people who insist on getting off busses via the front door, blocking the incomers. An easy small step to reduce dwell time would be to require everyone but those people who need the ramp to get off via the back door.

    1. I see quite a few people like to have a few friendly words to the driver as they hop off. Probably important for social health.

      1. I’ve seen a driver make a comment about people getting off at the front – along the lines of “you should get off at the back” – and he was given an earful by one of the people alighting.

        Bus layout design is a factor – the double-deckers feel much more natural to get off at the back, whereas some of the singledeckers make it awkward to make your way to the back doors, especially when busy.

        1. On the tiny little Link buses it is practically impossible to get off at the back, they are so jammed with people standing in the little aisle and in front of the single little back door… you’d miss your stop if you tried to exit the front of the bus from the back door.

  21. Matt is it possible using this analysis to determine what effect a city wide congestion road pricing scheme would have on bus speeds at peak times? And the flow on effect for increased peak time bus frequency?

    This would improve service. Which added to the improved cost relative to motor vehicles due to the congestion charge must improve patronage. More patronage might allow further increases in frequency, thus creating an upward spiral of improving PT and increasing patronage.

    1. Further benefits of a electronic GPS type congestion charging system is it could lower transaction costs for Councils policing car parking management systems. It may even be able to automatically fine motorists who infringe on bus lanes.

      The revenue generated could be spent on a mix of improving PT and decreasing fare costs (preferably by monthly/yearly passes) for the least well off groups.

  22. Great article, Matt. I think further bus prioritisation on key routes is also part of the solution for faster and more attractive bus services, ie further implementation of transit lanes, and reviewing existing T2s for a T3 upgrade, T3 to bus lane upgrade etc. Rigorous enforcement and stiffer fines would also help.

    There will be a backlash from motorists – I’m only too aware of that in the Northcote/Birkenhead area where local National MP Dan Bidois is campaigning against further transit lanes, and is seeking to water down the Onewa Rd T3 to a T2. Sigh.

    But we really do need to explore all avenues to make bus travel and PT in general as fast, efficient and safe as possible to encourage mode shift. So thanks for your informed analysis, and I hope AT get on board with this and use data, analysis and good policy to guide their decisions without political interference.

  23. Light rail (trams) work in Melbourne, and used to work in Auckland because they ran down the middle of the road and people hopped on and off at tram stop islands in the middle of the road. This could be replicated with buses, having bus stop islands at traffic lights – outside the old Dominion Road post office springs to mind – whereby passengers would cross to the island at the lights and wait for their bus. Having the bus lanes in the middle of the road instead of down the side, could help speed up bus flows.

Leave a Reply

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