Earlier this year I undertook a rather long and splendid journey starting in Amsterdam, the Netherlands and ending in Lisbon, Portugal. In seven previous posts I document our progress as follows:

  1. Amsterdam to Annecy
  2. Annecy to Cassis
  3. Cassis to Llanca
  4. Llanca to Zaragoza
  5. Zaragoza to San Sebastian
  6. San Sebastian to Gijon
  7. Gijon to Santiago de Compostela

The routes we took are also illustrated in the map below. At various stages in our journey we used different transport modes, including combination of plane, car, bicycle, train, bus, and ride-share.

At first glance this may seem like a strange route. After all, the natural line of travel from the south of France is to continue south along the east coast of Spain. Two seasonal factors influenced our decision to head west instead (NB: We were travelling in July and the end of August), namely 1) the heat of the Spanish summer and 2) the influx of summer tourists. Spain’s Atlantic coast is both milder (20 – 30 degrees) and receives fewer summer tourists. So if you do go to Spain in the summer, then I’d recommmed heading north-west.

In this post, I document the final leg of our journey, which us from Santiago de Compostela, Spain to Lisbon, Portugal. We travelled by BlaBlaCar to Porto, and then caught the train to Lisbon. Combining two transport modes enabled us to save us both time and money. More specifically, the train from Santiago de Compostela to Porto runs infrequently (every 3 hours), is slow (4.5 hours), and relatively expensive (35 Euro per person). In comparison, BlaBlaCar picked us up from our door, took 2.5 hours, and cost only 16 Eur per person.

An apt way to end our journey, I thought. One of my fellow bloggers (a’hem, PATRICK) expressed bemusement that we didn’t use the train to travel everywhere. I freely admit rail is nice, where it exists, is frequent, fast, and affordable. On our journey, however, not all these criteria were met all of the time. The two reasons I gave to Patrick for using non-train transpot modes, which I think is worth repeating here, are:

  1. If you want to travel off the beaten track in Europe, ***then*** sometimes you will need to make use of transport modes other than rail. This is especially true when travelling in countries with less well-developed rail networks; and
  2. If you are travelling in Europe at peak times, e.g. summer and/or oevr Christmas, ***then*** you can expect that the high-speed trains will be expensive if not completely full on some days.

For these two reasons I’d strongly suggest that people add buses and BlaBlacar to their list of back-up transport modes. When combined with rail, they make it really easy to travel in Europe without hiring your own car. Of course, neither of these modes is perfect either, hence the need to be flexible.

There, multi-modal sales pitch over. Having arrived in Lisbon, we then set about enjoying ourselves even more than we did on the train. Being late August, the temperature was pushing 30 degrees. I must say that we had 5 days in Lisbon, and it wasn’t enough. Two things strike you almost immediately about Lisbon: 1) the city is extremely beautiful and 2) the geography and topography is spectacular, albeit something that makes it harder to get around.

From a transport perspective, the most interesting aspect of Lisbon has to be the quaint trams that navigate through very narrow streets and up very steep hills, as shown below.

Apart from exploring the city, including the castle shown in the prevous image, we also took two day-trips further afield. One day we caught a suburban train along the coast to a sea-side village called Cascais. The train takes ~40 minutes, and runs right along the coast.

I can report that dwell-times on Lisbon’s (heavy) rail line to Cascais are approximately 25-35 seconds, even under peak summer loads. I mention this because Portugal is not, shall we say, the wealthiest and/or most technologically-advanced nation in Europe. Nonetheless it still manages to achieve dwell-times half that of Auckland. Shame on us.

If Phil Goff and the new Council need any convincing that they should push AT on the dwell-times, then I’d suggest we put them, first, on a plane to Lisbon and, second, on a train to Cascais. Perhaps we could even find a European watch-maker to sponsor the trip and supply watches so the Councillors can precisely time the dwell-times?

On our other day-trip, we rented a car and headed to the UNESCO world-heritage site of Sintra. While Lisbon is a wonderful place to visit in of itself, Sintra moves the region into the top-shelf. Sintra manages to combine both natural and historical beauty in a way I’ve not really experienced anywhere else, and which is as a result somewhat hard to explain.

I would say that Sintra is best understood as a sub-region, which is dominated by a large forest and mountain range within which are sprinkled an amazing number of amazingly beautiful places of interest, such as public gardens, villages, palaces, castles, and convents. Here’s a few images to whet the appetite (NB: All images are grabbed from the web; let me know if they are yours and you would like credit and/or them taken down). If you’re intrigued, then check out the Wiki page for more detail.

Sintra is only an hours’ drive from Lisbon, or there is also a train from Lisbon. Just be aware that the coast west of Sintra is also worth exploring, and cannot be reached by train.

Having spent five wonderful days in Lisbon, and four weeks travelling across Europe, we then flew back to Amsterdam so that I could begin the current academic year. If you are interested in the relative cost and speed of the different travel modes we used then let me know and I’ll try and write up a summary post. Otherwise, I have another upcoming travel post which considers our Christmas adventure in Bordeaux.

Until then, I will leave you to enjoy the New Zealand summer, of which I am suitably jealous. Travel safe y’all.

Share this


    1. There’s at least a couple of ways to improve dwell-times (NB: All of which are standard on rail operations overseas):
      1 Enable passengers to request the doors prior to the train stopping
      2 Get rid of the auto-extending platform (which can instead be extended upon request as/when necessary)
      3 Have the driver run the door-opening process, rather than the Train Managers (removes one person from the process)
      4 Generally just tighten up the processes

      Basically, just copy things that are standard-practice overseas. There may be others I have missed — I’m not an expert on these issues, but there’s no obvious reason why dwell-times on Auckland’s trains should be twice as long as good-practice systems overseas.

      1. I agree that the current EMU dwell times are excessive and that AT should be doing more to reduce them. However as any changes may entail both modifications to the EMU hardware and software as well as Transdev’s operator licence safety case as well as the possibility of inciting industrial action from the RMTU, AT may have put it in the too hard basket. This would be a pity as the impacts of long dwell times in overall network performance will only increase as CRL etc comes on line.

        Citing other rail systems as examples however does need to include the context that they operate in. For example the line in Lisbon that was quoted seems to have high platforms and level boarding (http://flickrhivemind.net/Tags/cascais,emu/Interesting) so the compromises on platform height that had to be made in Auckland to accomodate a variety of passenger and freight rolling stock may not apply there. Given that the extending step on the AT EMU centre car was an effort to improve boarding for wheelchairs and buggies etc, I’m not sure that converting it to NZ on request operation would be viewed favourably by various stakeholder groups. No doubt viable solutions could be devised but I expect the design , safety validation and implementation times and costs might be significant.

        1. hi Alphatron, thanks for the useful background information. I appreciate these things are complex, but also believe your statement that “AT may have put this in the too hard basket” is ultimately the reason why the dwell-time isue has persisted for so long.

          Running PT systems is difficult; AT need to come to terms with that fact. And they also need to come to terms with the fact that 50-60 second average dwell-time is unacceptaby inefficient. So while fixes will cost money/time, this is something that they can’t avoid doing — in my view.

        2. If AT continued to pay train managers etc. then there wouldn’t be industrial action. Sometimes it’s best to pay for people to do nothing

        3. I think the RMTU is smart enough to work out that a job doing nothing is much less secure than a job doing something.

      2. Stuart there is not really much time saving there.
        1. May save one second, two tops.
        2. Auto extending ramps are unfortunately required due to the design, CAF say they can’t be left part extended (the same distance as the fixed steps that were retro fitted themselves due to the distance we have our platforms from the rail). “can instead be extended upon request” by who once you have removed the TM as part of step 3?
        3. Driver only operations do not remove the TM as all but one state in Australia (so far only Victoria don’t have TMs but are apparently looking at returning them) has tried and then restored the position (many just doing doors from the back cab ie no customer service position).
        4. Tighten up the processes how?

        Dwell times on Auckland’s trains are due to the door system AT requested that they were told at the time was slow but someone in AT liked the cool look so that is what we got. The computer may be able to speed the door operation up but that will result in the possibility of its own safety issues, the slowest part of the door operation is the ramp that as explained in point 2 above must remain as per CAF.

        1. Bigted: re point 3, maybe not in Oz (for whatever reason), but driver-only operation does enable the removal of TMs. For instance, the entire London Underground and Overground networks don’t have any TMs or equivalent, with the driver having complete operational responsibility.

          Re point 2, on request of the passenger, for example by pressing a special button. Croydon’s latest trams have a button marked “Pram”, which give an extended door opening time and could similarly be used for ramps,

          But as Alphatron says, none of the possible options will be quick, easy or cheap.

        2. Auckland’s infrastructure or lack of it is far closer to the Australian railways than the UK so we should stop comparing our operations with that of anywhere in Europe until we have matching infrastructure.
          The ramp will always need to be deployed unless the loss of lives or limbs is an acceptable trade off to get the trains running 60 seconds faster (over a full trip).
          Correct none of the possible options will be quick (noticeably quicker operations), easy or cheap (retro fitting onto existing units to fix a major design flaw is neither easy or cheap).

        3. With Auckland’s ETCS and electrification infrastructure, rolling stock and operator all being European, it’s hard to see how technically it has more in common with Australian systems than European ones.

        4. Mike take out the actual rolling stock and our Auckland rail system wouldn’t even come close to a backwater regional system in Europe, as for ETCS where it the first world do they still use ETCS level 1?

        5. Bigted, most European operations similar to Auckland’s haven’t even got ETCS yet! In fact, Auckland is more European than the Europeans in that respect, no other city in the world (I think) having yet got to full ETCS implementation (whatever level) for mainline passenger trains.

        6. So why make Auckland the guinea pig? ETCS for all of its safety features plays a big part in slowing the Auckland trains considerably.

        7. Bull, BT. These ETCS trains are quite capable of performing significantly faster than the current timetable even as they are set up now, doors and buffers and all and they can be set up to go even faster.

        8. JC Jones even with artificially slow speeds being enforced by the ETCS when leaving platforms due to poorly located stop markers (in relation to the signal and balise), after passing a yellow (just about every light is yellow during peak) and in so many other situations?

          The reason AT is not making noise about speeding up doors etc is because everything is to the spec they bought (despite some advice against it).

        9. That’s complete rubbish Bigted. I can see two sets of track from my desk at work and both of them frequently have green signals during peak, how would a train get from Britomart to Ellerslie in 16 minutes if all of the signals were orange.

        10. Jezza the difference between a green and yellow is to do with the next signal, a yellow is as good as a green until you have to stop for a platform then the release speed is limited until the next signal while the ETCS wait to cross the balise to be sure it is at proceed.

        11. Not sure you are correct regarding Australia. From my observation there are no TMs in Adelaide or Perth either, of the five main cities, two have TMs and three don’t.

        12. Sydney and Brisbane both have TMs along with the rest of NSW and Queensland. They are door operators from the rear cab only with no customer service/interaction role.

        13. Agree that NSW and Queensland have them, but Victoria, South Australia and Western Australia don’t. Your statement that only Victoria doesn’t have TMs is not correct.

        14. Sorry, but wrong again, Bigted: according to the Sydney Trains Trainee Train Guards Information Pack, duties include providing information to customers and customer service as required. Their employer clearly thinks that they do have a customer service/interaction role.

          And your comment about Melbourne thinking of reintroducing guards is, I think, misleading. As I understand it they’re talking about security guards (perhaps even armed), nothing to do with train (or door) operation.

        15. Nice one Bigted, very droll, censors instead of sensors, hence shifting the problem focus from a mere technical issue to a censoring management refusal to allow improvement

        16. When you say the dwell times are due to the door system AT chose what do you mean? The only thing I can see that I haven’t seen on overseas systems is the extendable ramp, which does not explain anywhere near a 20-30 second difference in dwell times. Everything else about our door systems is also used overseas on networks with considerably faster dwell times.

        17. 20-30 second difference? There would be no comparable system anywhere in the world that has 10-15 second dwell times.

        18. That’s an optimistic view of our dwell times at the moment! I time them sometimes at Greenlane and Meadowbank, two stations with below average passenger volumes and the fastest I have ever clocked is 48 sec, typically it is between 50 and 54 sec from stop to start.

          I’ve travelled on a number of systems overseas with dwell times around 20 – 25 sec, which is where I get my 20 – 30 sec difference from.

        19. Over 30 of which is the computer firstly telling the doors the wheels have stopped so they can open followed by the doors individually reporting back to the computer they are closed before the 6 seconds the computer takes to release the drivers controls back to the driver, that time does not include anyone getting on or off. Around half of the door operation times are due solely to the retractable ramp so if there is no wheelchair, bicycle or pram passengers the door operation on the middle car is completely wasted time.

        20. Agree that a significant proportion of the delay is the computer communicating to the doors etc, but I’m not sure what this has to do with your often stated reason for long dwell times that ‘someone at AT chose the cool looking doors’. Are you talking about the ramps when you make that reference?

        21. Why does it take 30 seconds for one set of sensors to report to the computer, then for the computer to release the doors. That should take 30 milliseconds.

          The ramps should automatically deploy the second the driver releases the doors. If half of the door open time is deploying the ramp, then that is the critical path and should set the dwell time. If it takes 10 seconds to deploy the ramps and six seconds for the doors to report, then the ramps should be deployed and the doors opening 10 seconds after stop.

        22. Ok so why does it take 30 seconds for a door cycle? It only takes three seconds for the doors to physically open, so why does it take a computer 27 seconds to do the rest? My cellphone can compute a billion digits of Pi in that amount of time, what exactly takes so long?

        23. Nick do you remember Windows XP? That is what these trains run on and with multiple sensors per door all telling the main computer at the same time they are closed take a little time and as I’ve said before there is then 6 seconds once everything is set before it hands over controls back to the driver.

        24. Bigted – so you’re claiming that these trains run on software for which support ceased 3 weeks before the first services run using the electric trains in 2014, something I would highly doubt. This suggests you are just making this stuff up, as do a few other comments you have made in the past.

        25. Sorry I don’t buy it, even an older computer can process sensors like that in a few milliseconds. And why is there six seconds handover. What takes six thousand milliseconds to do?

        26. Bigted – while I don’t believe you for a minute that these trains use Windows XP as their operating system, even if they did there is no reason XP couldn’t play it’s part in processing signals from sensors in a few milliseconds. I’ve ridden on CAF units in Santiago where the doors open slightly before the train stops and I have no reason to doubt that these units are capable of that (whether it is something we want or not). All evidence points towards AT, Kiwirail and the unions needing to put there heads together and solve the problem rather than find reasons it can’t be done.

        27. Greenlane is a timed station, so will often have artificially long dwell times. Meadowbank has visibility problems that can also lengthen dwell times. Maybe try some other locations. I’ve timed many stops and have logged everything from 27 seconds up to 25 minutes. The “official” target range is 38-42 seconds and most crews can do that when they need to, but quite frankly, with this timetable they usually don’t need to.

          BT spends his day carrying a clipboard and gossiping with disgruntled drivers, so don’t fall for his play at technical credibility. The computers that run (the whole train, including) the doors are Linux based. The ETCS computer is Win-XP. The ramps add 4 seconds each way (if used). The zero-speed sensor takes up to 2 seconds, but usually less. There are good reasons for that, but it could be reduced if the consequent hazards could be mitigated or accepted. The other delays are built-in to increase reliability and ensure compliance. It’s not just the computer having a Friday afternoon chinwag with the doors. The lag is intentional. The doors move too slowly. The ramps even slower. These are all things that could be improved if the costs were reflected appropriately in the benefits, but it seems that for reasons which remain secret, they currently do not.

        28. JC – thanks for the technical information, it somewhat confirms what I suspected – it appears to mainly be to do with protocols rather than the technical capabilities of the trains. These are things that can generally be solved by getting heads together rather than requiring hugely expensive rebuilds.

          I’ll time some other stops when I remember, although I’m not sure Greenlane is a timed stop as you say, it’s certainly not listed as one in the timetable.

          Don’t worry I haven’t been fooled by Bigted’s technical claims – it’s usually reasonably obvious from his comments, although I do owe him an apology for questioning XP being used.

        29. JC Jones Greenlane is not a timed station but it is possible up trains would often be on a red there due to the small gap that southern trains follow Onehunga’s, the Onehunga waiting on Remuera (for Newmarket congestion) holding the southern train at Greenlane.

          jezza just a touch under 20 seconds for the wheelchair ramp to deploy, the door to open and then be immediately closed (without time for anyone to use the door) and retract the ramp. The computer still needs to do its thing pre green light and after all doors are closed, I don’t know that the process is able to be sped up much if at all safely.

        30. Well the benefits should be clear. If we can save twenty seconds per stop across the network, that’s approximately 30 vehicle-minutes saved per hour at peak. That equals three whole trains when running at 10 minute headways (three more units for longer trains) and something like 200 passenger-hours per hour of passenger benefits.

      3. 1 – That’s the smallest of the delays in the system. Not a significant dwell issue, but a trade off between user ease and fault/wear minimisation.
        2 – Yep, something has to change there. I believe there’s an issue with liability for fixing that mistake. No-one wants to pay for it.
        3 – The train manager doesn’t unlock or open the doors, the driver does. TM only closes the doors. Wellington manages fast dwell times with TMs operating under exactly the same rules and methods. It’s the equipment.
        4 – Already done and documented here last year. Amounted to only a few seconds and some of it appears to short circuit safe working practices.

        1. Note they were just suggestions – I’m sure there’s other things that can be done. After all, we can shave 30 seconds off these dwell-times and still only be on par with what is normal overseas.

          Also, as I indicated above I am not an expert on rail dwell-times, but I do know that Auckland’s are too long. The point of the post was not to discuss ways to shorten dwell-times (even if the comment thread has headed in that direction), but simply to note that Auckland’s need to be shorter.

    1. perhaps. However I don’t think it’s an issue with our lack of technical knowledge, but rather a lack of leadership and oversight at AT that allowed long dwell-times to become an issue and persist for so long.

      1. I think it may be a lack of technical knowledge. If it were a technical fix then the people ca[able of the fix would/should have passed this info up to the management/leadership for action. Feedback is necessary to guide and teach managers and leaders
        Maybe Mr Goff needs to ascertain if it is a technical knowledge issue or an incompetent/feet dragging management and then take action to resolve the problem

        1. I suggest that you read Alphatron’s comment above; consider the number of bodies that need to be involved, each with their own particular stake, history of onvolvement and agenda (eg AC, AT, NZTA, KiwiRail, CAF, RMTU, Transdev); recognise the heavily-regulated nature of the rail industry; and then reconsider whether the issues (for there are many) can be explained by a simple black-and-white question of technicality or competence.

          Unfortunately real life is not as straightforward as that, as I’m sure someone with Mr Goff’s experience would appreciate. I would imagine that there is lots of work going on behind the scenes looking at the significance of each of the many issues and how they inter-relate in the complex railway setup – and, of course, who pays.

          Whatever the full issues may turn out to be, neither what they do in the very different Lisbon environment nor kicking butt are likely to produce the optimum result.

        2. hi mike, thanks for your comment. I just had a couple of thoughts in response:
          1 I appreciate the dwell-time issue is complex,
          2 but it has to be fixed. Period. There’s no way Auckland’s rail system can function efficiently in the long run with dwell–times of 50-60 seconds per stop. As others have noted, this adds 5-10 minutes to each and every run, and means that average speeds are slower than steam engines.
          3 from what we see externally, there is a general impression that AT do not take this issue seriously enough. Hence I think there is value in external observers, such as the blog, regularly noting it is an issue — where relevant.
          4 I’ve heard from people who have studied dwell-times in Auckland that it should be relatively straightforward to trim 15-20 seconds of Auckland’s dwell-times with some small to medium sized fixes. I’m not exactly sure of what this involves, but it would be a significant gain if it could be achieved.

        3. Somewhere in the middle of all that lot is Transdev who I thought were an international operator with experience in best practise. With so many organisational stakeholders it’s no wonder there isn’t a meaningful customer service focus.
          What is it with New Zealand with railway inefficiency? Wasn’t this part of the reason for the ’80s privatisation?

          I would be happy if Phil Goff’s legacy was streamlining some of this lot and reducing dwell times. Aim high!

        4. Yes i think Transdev is a complication but not necessarily a barrier to change. AT really needs to drive this.

        5. Mike – working in a sizeable bureaucracy I would agree with you that there is a significant amount of complexity, silo thinking and unwillingness to change. However, from experience it is amazing how quickly this changes when some heat is applied from above, something I suspect has been missing. Hopefully Phil will achieve this, although of course he is not in charge of all of the stakeholders involved.

        6. stu: “it has to be fixed”. Does it? Do the benefits exceed the costs? I don’t think we know enough about the current situation to answer that (post CRL is a different matter).

          JT: “What is it with New Zealand with railway inefficiency? Wasn’t this part of the reason for the ’80s [actually 90s] privatisation?” – it was that overall process that helped create the plethora of organisations involved nowadays!

          One area that hasn’t been touched on as far as I’m aware is whether the trains are performing to specification with respect to dwell times and door processes. If they are, Mr Goff can apply all the heat he likes but AC/AT will be paying the bill; if they’re not, he should be banging on CAF’s door; if it’s unclear, bring in the lawyers…

        7. Mike – in short it does have to be fixed. Dwell times of 50 sec at quiet stations will likely mean well in excess of 60 sec at Aotea and Britomart once the CRL opens, which is not viable with the frequencies planned, meaning it won’t be possible to get full value from a $2 – 3 billion investment.

          My understanding is that the contract stipulated dwell times of 30 – 40 sec with TMs in place, with the ability to go driver only in the future further reducing the times. If this is correct then CAF are definitely in breach, unless it is AT/Kiwirail cautiousness that is adding to the delays.

        8. Jezza at these 50 second dwell time stations does the driver have a proceed signal or is the signal the reason for the long stop and nothing else?
          Drivers will tell you they haven’t seen many green lights for a long time, lots yellows (indicating the next is currently red) and reds so the signalling may be causing some of the issues you blame on the TMs.

          You are dreaming if you think DOO is going to have shorter dwell times than are possible under the current TM system. ADLs will piss all over the EMUs in regards to dwell times due to the superior (user friendly) door system they have that is fully TM controlled. ADLs would even give the EMUs a run for their money in overall speed (even though the top allowable speed of an ADL in 20kph slower) as they aren’t slowed by the over zealous ETCS system that constantly hampers the EMUs.

        9. Hi Mike,

          You raise a good point about benefits/costs. Let’s try and work it out quickly:

          – Operational benefits of achieving normal practice well-times = 30 second saving per station. This would save 5-10 minutes on every trip from Swanson/Papakura to City, which is about 10-15%. For sake of argument, let’s assume this results in a 5% savings on total operational costs of $200 million p.a. = $10 million p.a., probably growing over time as service levels increase with CRL etc. NPV of this saving over 30 years is likely to be in the order of $150-$200 million.
          – User benefits from faster travel times: Let’s assume there are an average of 30 million passengers per year over the next 30 years, who travel an average of 15km per journey. Average station spacing is say one per every 3km, or 3 intermediate stations per journey. Saving 30 seconds per station equates to to 1.5 minutes per journey. Total value is then (1.5/60)hours per journey x $10 per hour x 30 million journeys per year = $7.5 million p.a. NPV of this is in the range of $112.5 – $150 million.

          Unless I’ve made a mistake somewhere (quite possible) these two benefits (operating costs and user benefits) alone give you $262 – $350 million in benefits over 30 years.

          That’s before calculating the congestion reduction benefits associated with diversion of car trips.

          Given that total cost of EMUs was in the order of $500 million that suggests it’s worthwhile fixing this issue, even if we have to basically re-build the EMUs.

        10. Bigted – of course DOO will have shorter dwell times than the current TM system, for the simple reason that with DOO all doors can be closed at once, whereas with the current system all other doors are closed, then the TMs door is closed.

          I agree the standard dwell times with the ADLs were quite a bit quicker, however the doors were narrower and also the train was often moving with the TMs door still closing, which I’m not sure would meet safety protocols. It’s interesting you call the ADLs door system ‘superior’ when it required a ramp to be carried into place every time a wheelchair passenger needed to get on or off.

          It’s odd you say the drivers hardly see any green signals, mostly orange and red, yet they are generally proceeding at full speed once they depart these stations from my observation.

        11. jezza they proceed as fast as the ETCS allows them to. The ADL door system is generally superior with the exception of the wheelchair ramp (something a good TM can deploy nearly as quick as the EMUs ramps deploy but are only required when wheelchair passengers are riding) and that is why the new trains in Wellington use what became known as the ‘Auckland door system’, AT on the other hand was more interested in looks than operation and are now stuck with exactly what they paid for.

        12. Bigted – are you saying the ETCS allows them to proceed at full speed even if there is an orange light?

          You have made this claim numerous times about AT wanting the doors with looks over functionality, but any time I ask you to explain what you mean you quickly go quiet, can you explain what you mean with this assertion?

        13. Jezza they had the option to have the tried and tested (as WRC did when ordering their trains) but whether it was over excitement or whatever someone opted for looks over functionality.

        14. What part of the look of the doors makes them slower? If they move slower than the old DMU doors that sounds like a setting to me.

        15. They are set as per ATs specs, changing the specs requires things that AT don’t seem bothered with at the moment.

        16. Right so it’s a problem with the operating settings not the design. That’s great because that sounds much easier than having to change the design, you appeared to be giving the impression that the design was poor.

        17. They went for the cool look rather than worry about how it operated and how the specs they ordered effected everything else.

        18. The door design appears to be reasonably common overseas, it’s used in both Melbourne and Sydney, both of which have very quick dwell times. I’m not sure it plays much of a role in the dwell times, maybe the way AT chooses to operate them does, but as I say that is at least configurable so can be solved at some point.

          I’m getting the feeling a lot of the reasons behind the slow dwell times are configuration and rules rather than design, which leaves me optimistic that we will be able to get the dwell times down at some point in the future.

  1. Had a TM that was a bit quicker with the door close which seemed to reduce the dwell time by 10 secs. Maybe less time standing on the platform blowing whistles might help. Driver seems to be waiting for the TM at Newmarket quite often. Unless that’s one of these time stations.

    1. Newmarket is a timed station at all times except during AM peak (0700-0900) according to the timetable. Newmarket will one of the stations that TMs are often waiting for a signal into Britomart due to congestion.

      1. Do you mean driver waiting for a signal? So the journey is delayed while the TM’s wait until the scheduled departure time at Newmarket and then is further delayed by the signals.

        Do they make Newmarket non-timed in the afternoon peak as well as the morning?

Leave a Reply

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