This is part six of my series looking into how we can make big improvements to bus journeys through relatively small changes.

  • In part one I looked at Khyber Pass and the need for us to be more flexible in providing better bus lanes.
  • In part two I looked at how better bus priority measures can help improve the quality of our town centres.
  • In part three I looked at Great South Road between Newmarket and Ellerslie
  • In part four I looked at some of the changes we need to make to Link Bus services
  • In part five I looked at some issues with the process for setting bus and transit lanes

For part six of these series we are going to look at how to improve dwell times as well as discuss whether we are managing buses right.

Improving Dwell Times

Some of the ways we can make Auckland bus services better is by reducing dwells at stations either by speeding up boarding or alighting.

Dwell times have already improved a lot in recent years, thanks to the very high use of the HOP card. With no need for slow driver interaction, this has reduced dwells immensely. It’s hard to imagine the mess we would have with today’s ridership numbers with everyone queuing up to have the driver individually press a series of buttons to issue a ticket.

Still, more can be done and we should continue to learn from what works well overseas, such as in Melbourne – which I wrote about last year. In Melbourne it is so much easier to buy a Myki compared to buying a HOP card (I know Myki has immense flaws such as cost amongst other issues, but the accessibility of getting a card isn’t one of them). The easier it is for a tourists, first time & occasional users to acquire a card, the closer we will get to 100% usage of HOP, and may even mean we can look at cashless in some locations (especially the city centre) in the future.

Another way we can improve dwell times is through all door boarding which is in use for the Northern Express – but weirdly for no other bus routes. AT should really consider this at busy major stops.

The Webster Ave Trial in New York City led to impressive improvements, according to NACTO:

Comparing service from a year before installation to a year after, bus travel times through the corridor dropped 19% to 23% for rapid buses. A Bx41 SBS trip during the PM peak fell to 40 minutes, compared to 52 minutes on the previously operated Bx41 Limited. The local bus also saw benefits, with trip times reduced by 11 to 17%

Similar results were seen in San Francisco

San Francisco shows the specific benefit of all-door boarding. At busy stops, a 38% reduction in entry/exit time was found for buses: 1.5 seconds per customer, in a system with 100 boardings per bus in the peak hour. Transit travel speeds increased 2% on average after implementation. Coupled with improved enforcement, fare evasion dropped from 9.5% to 7.9%, reducing estimated fare loss nearly $2 million.

Melbourne Tram All Door Boarding Readers

Another way is to make sure all new buses definitely consider dwell times as part of the design for example around doors. More education around double deckers could help too. Don’t be that person who gets on at Akoranga, goes to the top deck and gets off Fanshawe. Don’t be that person, just don’t.

Smarter Use of Buses

One of the most frustrating things I see on the network is uneven loadings, for example, seeing an empty bus leave straight after a full one on a busy route. You will often see this in the evening peak with the Dominion Rd buses where a late bus pull in, fill up to standing room only then straight after another bus will pull up. Of course, with everyone waiting having packed onto the first bus and the second bus timetabled to go it leaves at the same time or very shortly after with hardly any people on it. When you get over a certain buses per hour, keeping rigidly to the timetable becomes rather silly and actually counter productive.

It seems on busy bus corridors AT has just kept adding more buses, rather than more actively managing the dispatch of services to encourage more even loading of buses. Adding extra peak time services is really expensive (you need a new vehicle and driver), so there are some huge operational efficiencies to be found through getting smarter about bus operations instead of just throwing peak capacity at the problem all the time.

I often also used to see this on Mt Eden Rd in the mornings citybound, where you will see a single decker bus nearly full with an empty double-decker bus bunching straight behind.

I’m sure there are plenty of ways that this is managed overseas, please comment below if you have any ideas on how to better manage these buses or know about any ideas used overseas on busy bus corridors.

Some of the technology identified in the Advanced Bus Solution, which we rightly ridiculed as a poorly developed alternative to Dominion Road LRT, might actually be useful for squeezing more out of some of our other busy busy corridors. These slides are taken from the publicly released reports I read at the time.

Advance Bus Solution Appendix Page 23
Advance Bus Solution Appendix Page 90
Advance Bus Solution Appendix Page 91

In summary:

  • buses can be linked to the traffic light system to reduce the time they spend waiting at red lights
  • less focus on strict timetables but ensuring buses are always coming a few minutes apart
  • a high degree of monitoring of routes, which can be done with either data from the previous weeks operations, or even live from a control centre.

Where AT are doing a good job (and hopefully we will see more of this) is the move to higher capacity buses, whether this is double deckers or newer triple axle single decker buses replacing those really inefficient small double axle buses.

I hope AT do study into how future fleet procurement can result in even higher capacity buses, for example I would be interested to know if higher capacity metro-style buses for inner isthmus bus routes with less seating and more standing room could work, like Matt has proposed we could do for our trains.

Another possibility is looking into the potential for articulated buses that common in Europe and proposed in Advanced Bus Study for routes which are unlikely to need Light Rail in the foreseeable future, or earlier procurement of electric buses than the agreement Phil Goff has signed for procurement of only zero-emission buses from 2025.

Van Hool Exqui City Bus

Finally, a key issue that drives down the efficiency of our bus system are “padded timetables”, which is when the timetable is based on a “worst case scenario” for a route’s running time – meaning the bus or train needs to travel much slower than it should. There’s no doubt Auckland Transport uses massively padded timetables to help improve service punctuality performance, especially on the rail network that continues to run slower than when we ran clapped-out old diesel trains. For buses, padded timetables are a bit more necessary because routes without bus lanes are subject to the highly variable impacts of congestion. Nevertheless, padded timetables are terrible as not only do they create slower journeys, they also mean we are wasting lots of OPEX running more buses per hour on a route than we need to. The other day I saw a route 18 bus laying over behind the next 18, this means that bus with its driver would have been laying over for more than 15 minutes waiting which is a complete waste of money.

Implementing all-door boarding and managing our buses better could for a very low cost, pay huge dividends in both cost savings and travel time improvements.

Share this


  1. I agree, i think overall dwell times have got a lot better in the last few years. For me the biggest problem is congestion and timetable padding. i hate being on a bus, when the driver clearly going slow, just to keep to timetable. But i also get how frustrating it is when bus turns up early to a stop.

    Also During peak times, the buses bunching together can sometimes just be caused by congestion, especially around central Auckland

  2. How about paying for a bus BEFORE getting on it, while waiting for it. Works for trains. Why not for buses?

      1. Not really, unless you show the driver your ticket/receipt as you are boarding but then thats probably slower. Otherwise you will bring rails issues with fare evasion to the bus system.

        Also not to mention VRDs are rather expensive to roll out to bus stops.

        1. The question then becomes ‘Are we happy to accept some level of fare evasion in exchange for a faster bus journey?’. I’d definitely be happy with (say 5%) fare evasion, 25% not so happy.

          We can also ask the question ‘can we reduce fare evasion while keeping all door boarding and off vehicle payment?’. Yes; we can employ transit police, we can gate stations (northern, northwestern, and south eastern busways).

          There is already fare evasion on the bus network, off board payment makes it easier, but do the benefits outweigh the costs or can we develop a system where they do?

        2. This is a good idea, and only needs to be done at the busiest stops at peak times. It’s pretty simple, either pay with hop or pre purchace a ticket with cash from a TVM on the footpath.

          Then you tag on or display the ticket as you board. If it’s all door boarding you have a stop marshal viewing the back door, which is exactly what they do already with the NEX.

          There is probably only five or six really busy stops where you’d do this routinely, no big deal.

        3. Is the driver going to stand at the back door if there is all door boarding and watch me tap on?

        4. No, but a fare inspector might. As long as the fine is not worth the risk.
          If all door boarding and no cash sales reduced journey times by 10%, the savings would probably pay for a fare inspector on 20% of services or more.

        5. The Sydney system is that all door boarding is only for peak times at specific stops, when a marshal is in attendance.

        6. It can’t be that difficult. I remember reading, quite a few years ago, how they did it in Curitiba, Brazil. I think they enclosed passengers who had pre-paid in glass tubes. No doubt expensive, but worth considering at high-usage stops.

        7. “they enclosed passengers who had pre-paid in glass tubes”…and fired them into the bus? I can see how that would be fast if the bus doesn’t need to stop. As long as the passengers stop bouncing around in time to buzz for their stop

        8. They waited in the tube, then boarded rapidly when the bus stopped in alignment with the tube.

    1. This used to happen in Christchurch in the 1960’s at peak hours there would be someone at the bus stops in the square selling tickets. You just showed your ticket to the driver when you boarded the bus.Today of course you could just have a ticket machine to sell the ticket but it will be a long long time before we have driverless buses if ever. It could even be compulsory to have a ticket at inner city stops. But the hop cards work well on buses and there cheaper but the have caused massive security problems on the trains.

  3. With bus bunching sometimes the following bus needs to let the first bus know it is empty and to pass it. The first bus with more people on it will fall behind anyway. But yes they should all be managed better – buses should be operating as fast as possible (and comfortable) to make the most of this asset (and to keep journey times down). Dwell times have definitely reduced with increasing HOP usage. Wouldn’t be hard to put some placards on DD advising people to use the upper deck for longer journeys and the lower deck for short trips.

    When are we going to get articulated double decker buses? – kidding/not kidding.

    1. NEX drivers do talk to each other and where one bus is fuller, will tell them to bypass a station, however more often than not a passenger will ring the bell and want to alight

      if the first bus stops at the back of the platform, then all the waiting passengers will start running back to it as the second bus overtakes to pull in at the front of the platform; it can all get a bit messy

  4. 10 years ago in London I used the 207 route in Acton quite a bit. They were articulated buses with all door boarding, huge doors, no cash payments (simple cash machines at each stop), sideways seats and tons of standing room. There was no timetable, they were every 3 minutes or so. Every so often you would be on a bus and an announcement would say the bus was being held at this stop to even out the flow.
    This is the exact kind of service we need on busy routes. And it would have to cut some serious money off the operation budget too.

  5. At Northcote, where I often get on the citybound 922, it’s not unusual for three or more passengers to get off via the front door, holding up a dozen or more passengers waiting to board. This is unacceptable, but how does one protest when the offenders are all elderly Asians who may not comprehend English, and we’re all too polite.

    On the subject of padding of timetables resulting in slow trips, the OuterLink is a notorious case. Westbound, it leaves the city and then stops at Victoria Park for 5 minutes or more while a driver change takes place. If you’ve ever endured a wait for this reason, you’ll know it’s not as simple as the new driver sliding behind the wheel and pulling out. The first driver has to log off, pack his/her belongings, the next driver gets on and logs in, steps out again 2-3 times to adjust the left-side mirror, adjusts the seat, adjusts the inside mirror, does several other things that I haven’t identified, and finally pulls out.

    1. Wide front enterance 2x tag machines with left in and left out would help solve that. I also notice back door is often only opened by driver if u head towards it so often delayed a bit so I will tend towards the front without good reason.

    2. We have video displays on buses, why not have slides in the most common languages, educating people – Use rear door.

      We only need four languages: English, Chinese, Korean, Hindi. We could add other languages as required when population sizes or tourist numbers increase.

      Just slip the slides in between others.

      1. It’s a good idea, but fares will rise to pay for it. I was also thinking about writing to Chinese etc newspapers and asking them to run an article.

  6. Articulated buses would make sense for busy routes that don’t go through the CBD. These don’t really exist at the moment however once the Eastern Busway and Airport to Puhinui connection are up and running this will change.

    1. Artics can also make sense if your CBD route is specifically designed to accommodate them. I think this will be possible for the NEX1 and NEX2 once we have implemented the new network.

    1. it’s an SUV and looks like one of the random small tour operators that use the busway, there’s room for them and they cause few issues unless they dawdle

      1. It’s a bus. A scheduled service capable of carrying more than 10 people. Usually they are airport shuttles.

  7. The only solution is to provide bus priority & bus lanes as needed where the PT level of service has dropped significantly (padded timetables, full/empty adjacent buses etc), and use simulation modelling to provide the evidence to justify the requirements.

  8. The bus knows how many people are on board (tag on/offs), why not just evaluate that on the central computer and message the full bus to miss stops (if no alighting) when there is a empty bus behind?

    Oh that’s right, AT bought a crappy French system, that can barely top up a card reliably, let lone communicate in real time!

  9. Fumble around in your bag for your hop card for 30 seconds. Based on my experience, 20% of people want to be that guy

  10. We should just replace the timetable for frequent bus service into minimum frequency format.

    For example peak time bus runs every 5-7 minutes.

    This remove the incentives to slow down the bus just to pad the timetable. Same concept can be apply to train once we got frequent train every 5 minutes.

  11. Yes surely with HOP data and GPS technology & driver displays we can have full or near full buses skip stops (unless someone wants off) & other trailing one pick up instead. Or alternatively close trailing emptyish bus overtake possibly by front one waiting artificially for 30’s at a stop or random stop location.

    On less frequent bus routes early buses are painful with padded timetables. Surely there should be a display for driver to slow down or speed up with minutes in arrears or advanced showing. If ahead of schedule the driver can take more time to help passages needing directions. If behind well perhaps not if they have a bus full of passengers. I’ve missed transfers because of overly helpful drivers when we are late to start with. Was that helpful to me? If early I don’t mind and have often given advice myself.

  12. I don’t see the logic in trying to prevent people getting on at Akoranga, using the top deck, and then getting off at Fanshawe.

    * If the’re getting on at Akoranga and getting off at Newmarket, they’ll take just as long to climb the stairs as they would if they were getting off at Fanshawe.

    * If they were getting off at Fanshawe, they’d take just as long to come down the stairs if they had got on at Albany as they would if they had got on at Akoranga.

    Delays boarding at Akoranga or alighting at Fanshawe are the same no matter where people subsequently alight or have previously boarded. The only way to keep these delays to a minimum is to ensure that no one uses the top deck. At all.

    Or have I missed a critical piece of logic?

    1. Often only one person will get on at Fanshawe or Victoria, so people boarding to the top deck hold up the entire bus.

    2. Yes and if you get off or on at the end of a route line it doesn’t usually matter as it’s going to sit there for a while & everyone else has to get off anyway.
      Basic etiquette is longer trip = OK to go to top deck, shorter trip = stick to the bottom if possible.

Leave a Reply

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