Back in August Auckland Transport kicked Snapper out of helping to deliver integrated ticketing in Auckland. Subject to any legal battles over whether compensation is due or not, this was the final line in a very sad and sorry tale about Snapper’s involvement in Auckland’s integrated ticketing system. A tale that goes back quite a few years.

The NZ Herald, and ourselves, have recently received a huge amount of information on the whole Snapper/HOP debacle – released under the Official Information Act. It may take a while to dig through the details of this all, but we shall give this task a go – because a huge amount of public money has gone into integrated ticketing plus a huge amount of time seems to have passed while we debate the whole issue. The Auditor General’s investigation into this whole shemozzle, which should provide us with some final answers around the whole sorry saga, is apparently on hold until Snapper’s legal battles with Auckland Transport have been resolved. So for now let’s just take a look through what we have – in particular looking at these two questions:

  • What were the reasons for Snapper originally not being chosen as the preferred tenderer for Auckland’s integrated ticketing solution?
  • What pressure, if any, did the government end up putting on Auckland Transport to allow Snapper to launch the Snapper/HOP (from now on referred to as SNOP) card early last year?

We’ll see what other interesting stuff we come across on the way.

A useful place to start is an NZTA Ministerial Briefing from September last year which provides some background information on integrated ticketing generally, but in particular the issue of branding and the inter-relationship between the SNOP card and other, future (at that point) HOP cards. Helpfully, the briefing summarises some of the technical details of the Thales system and how it differs from the Snapper system – so from the start we can begin to get an understanding of the potential difficulties in integrating the two:

From reading a bit more detail in other documents it seems that DESfire is what’s typically used for public transport based smart-cards. It does what is a reasonably simple job really really well and really really quickly. The Java JCOP is more commonly used for contactless cards purchasing small items. Like what Snapper does in terms of micro-retail payments.

Looking at another, more detailed, document – a presentation to the NZTA board – this distinction is highlighted further. The Thales system is designed around providing a really good solution for public transport and provides very fast read times by not complicating things too much.  The presentation, from 2009 and as part of the process of informing and educating the board about integrated ticketing before NZTA signed off on their pretty significant financial contribution towards the project, noted that the Thales system was of excellent quality and didn’t raise any concerns about ARTA’s choice of them as the preferred supplier:

The presentation has quite a lot of discussion around the extent to which third party equipment should be allowed in the system. There are a range of ways in which third-party equipment can form part of the integrated ticketing system – from different machines, different cards right through to separate systems and ‘clearing houses’. NZTA’s preference was to have as little third party equipment as possible, because this made the system a lot cheaper and a lot simpler.

So to answer our first question, it seems like Thales was probably chosen as the preferred supplier because their technology delivered better on the core goal of the integrated ticketing project: a smartcard for public transport purposes. Snapper’s different technology was more based around micro-retail. My experience with my SNOP card is certainly that it feels really really slow in using the card to have it register properly. I can put the card in front of the reader while still walking and find that it only registers when I’m just about to get beyond easy arm’s reach. With public transport efficiency so intertwined with reducing dwell times, slow readers are something you want to avoid.

The second question is perhaps the most politically sensitive, as touched upon in an NZ Herald article yesterday – the question of how much involvement central government might have had in persuading Auckland Transport to allow Snapper to become involved in the system even after they had lost the tender to Thales for being the preferred supplier. Here are the key points from yesterday’s article:

Labour transport spokesman Phil Twyford says correspondence between Wellington smart-card supplier Snapper and the Beehive shows the Government forced the former Auckland Regional Transport Authority to let the company join the Hop ticketing scheme after awarding the main contract to its rival, Thales.

Auckland Transport has since dumped Snapper for allegedly missing performance deadlines, a claim which the company denies.

Mr Twyford says Government denials of political interference are contradicted by a letter from Snapper chief executive Miki Szikszai to former Transport Minister Steven Joyce after a meeting in 2010.

Mr Szikszai wrote that he understood Mr Joyce’s “expectations are that … NZ Bus should be free … to implement Snapper equipment and join the Snapper scheme in Auckland.”

Here’s the letter that Mr Twyford is referring to. With the key extracts below: Perhaps reinforcing the likelihood of Steven Joyce wanting to get Snapper involved (perhaps hoping that the Thales solution would fail and Snapper would be able to take over the whole thing?) are a series of questions Joyce asked back in late 2008, when he had first become Minister of Transport and was obviously just coming to grips with the project for the first time. His questions were:

  • Why do we need a nationally transferable integrated ticketing system?
  • Why do we need to spend $100 million on an integrated ticket system for Auckland when Snapper has been provided for free and privately?
  • Can this project be stopped?

I think we obviously need to give Ministers a bit of leniency in their first couple of months – as without knowing the details these seem like reasonable questions to ask. But they certainly do give us the impression that Joyce was favourable towards Snapper and unfavourable towards the ARTA preferred Thales solution right at the start of his time as Minister. Fortunately the Ministry of Transport recommended strongly against stopping the project – and obviously that never happened.

I’m still working my way through the documents, which are in chronological order, and I imagine there might be quite a lot of more interesting stuff in recent times when it became clear that Snapper was failing to get things working with the Thales system – leading to them being cut from the project. But that will have to be the subject of another blog post.

Share this

52 comments

  1. I don’t think we need to get hung up on which card might have been better in discussion here. The fact is Thales won the contract for AIFS. That decision was the subject of a probity review at the request of Snapper. The review found nothing untoward with the process. Snapper then ask Steven Joyce and within days Snapper are allowed to participate in the AIFS project. This is banana republic stuff and should not be tolerated by John Key if his Government is to have any credibility.

  2. Agreed that the offered product is not the discussion, rather the process and contract arrangements that were in place. However I would not be prepared to lay blame with government or council until reading through the findings to see how snapper were let into the project. —- I’m not defending either of them —-

    However what we do know so far is that snapper was the problem, it’s just a question of who let them be the problem?

  3. “This is banana republic stuff and should not be tolerated by John Key if his Government is to have any credibility.”

    Key and the Government don’t care about credibility.

    National’s philosophy appears to be “We are the National Party and what we do will always be OK”.

  4. The SNOP card is definitely slow to be read. I have my new HOP card but can’t use it on the bus yet. Has anyone here used it on the train yet? Is it faster than SNOP?

  5. The fact that Snapper were lobbying politicians after the tender process was complete is almost scandalous.

    And the fact that Joyce, on the basis of this lobbying, felt compelled to ask “can this project be stopped?” is also condemning – because it indicates that Joyce’s opinion was influenced before he had even heard the arguments for the original decision.

    Does this not show that the Minister’s thinking was 1) too easily influenced by external lobbying and 2) that he spoke “mistruths” when he said that he had not exerted any pressure on the process?

    This is shaping up to be not very pretty indeed. While it’s too early to blame anyone, it seems fairly clear already that neither Snapper nor the Government are going to come out looking good. The only question is how bad?

    1. Prediction:
      1) Phil Twyford will ask questions in Parliament and the answers will be batted away. This line of response already started with Mr Joyce’s office saying Joyce “was leaving any response on the issue to Mr Brownlee, as the minister now responsible for transport.”
      2) John Key will claim full confidence in Steven Joyce. The letter incriminating Steven Joyce is merely someone else’s recollection. Due to his principle of hear no evil, see no evil, he won’t personally delve into it further as it will raise too many questions about Joyce.
      3) The Office of the Auditor General has already said that their investigation is on hold now that the issue is before the courts, so there won’t be any investigation from them.
      4) Snapper and Auckland Council will settle out of court, so that the whole sorry mess doesn’t become public and reflect badly on the both of them. The Office of the OAG won’t restart their investigation.
      5) Any other type of investigation won’t have teeth and will result in a whitewash.
      6) Voters will have the wool pulled over their eyes amid claims the whole thing is politically motivated by Labour.

      Have I missed anything? Jeez I’ve become cynical.

      1. Prediction 1 about to be played out today:

        PHIL TWYFORD to the Minister of Transport: Did Hon Steven Joyce meet with the Snapper CEO in March 2010 and discuss Snapper’s desire to enter the Auckland public transport integrated ticketing scheme; if so, why did he not include this in his answer to Oral Question No 10 on 27 June 2012?

        1. The line just taken by Steven Joyce in Parliament is that he directed NZTA to promote an open ticketing system. If that is the case why did NZTA not invite other prospective transport card providers to participate?

      2. You would expect the political interference line from Mike Lee but I have heard of other fairly important players who have said the same thing when they could have said nothing at all.

  6. “Why do we need a nationally transferable integrated ticketing system?”

    That’d be obvious really. So you only need the one card to catch the train, bus or ferry all around the country. A lot of us provincials catch the odd bus in Auckland sometimes, and yet we catch the bus in Welly, and we catch the bus in Palmy sometimes too, and in Welly there is the crazy system where half the buses (in the Northern suburbs, Porirua and Kapiti) have a different system to the rest of Welly and the Hutt. It’d be nice to have only the one smartcard to carry in your wallet for everywhere you go. Arrive in a town you don’t normally go to and you’ve already got your bus tickets organised.

    “Why do we need to spend $100 million on an integrated ticket system for Auckland when Snapper has been provided for free and privately?”

    A better question would be why do we need to spend $100 million on an integrated ticket system for Auckland when an open source solution exits – http://www.calypsonet-asso.org/ or see which cities are using it -http://www.calypsonet-asso.org/images/ENG_12_MAP%20copie.pdf I bet all of those cities didn’t pay $100 million dollars.

    In other news Adelaide’s system which was trialled quite successfully, has been launched today for everybody:
    http://www.adelaidenow.com.au/news/south-australia/adelaide-metro-launches-public-transport-smart-card/story-e6frea83-1226510509857
    Adelaide has the advantage of a time based system (2 hours with free transfers, doesn’t matter how far you ride) so you only have to tag on, and not off. Something that Auckland and Wellington should look at doing.

    1. You dont even need to look that far afield. Christchurch has a the same system as Adelaide and it always worked very well on the buses there. The two hour free transfer is great.

      The other thing that is good with the Chch system is the cumulative pricing. If you use the card more than twice in a day, the rest of the day is free. If you use it more than 10 times in a week the rest of the week is free. Basically meaning off peak is “free” to commuters.

      I dont understand why Auckland’s search for a system never seemed to mention the Chch system. Was it not scaleable? Or is it just a case of no good idea could come from anywhere south of the Cook Strait?

      Actually, considering the Minister of Transport’s record that may not be such a silly opinion to hold.

      1. I dont understand why Auckland’s search for a system never seemed to mention the Chch system. Was it not scaleable? Or is it just a case of no good idea could come from anywhere south of the Cook Strait?

        No, I think you’ll find it was from simple ignorance. The people making the decisions didn’t know of all the systems available and all of the systems available didn’t know of the tendering process with the end result being far more expensive than it should have been.

        1. That’s simply incorrect. The people making the decisions knew full well about Chch’s system (and other cities in NZ) before they even put it out to tender. As for why that technology provider did not submit/win I can only speculate, but my guess is that their technology is dated.

          P.s. Tag off data is really really important to designing a good system. Yes it slows things down a little, but the insight you get into how people use the system makes it so much better to design and deliver a good system.

  7. Cards cost less than a couple of dollars each (custom printed and all) in the 100,000 quantities, readers are < $1,000 each installed, around a 2,000 would cover everything well. A bunch of paper ticket terminals, say $10,000 each. A customer balance web site $1 mil, a backoffice consolidation and reporting system $2 mil (this is basic accounting, not rocket science). Rounded up for marketing that's only $10 mil, where did the other $90 mil go?

    1. Just off the top of my head – planning the whole thing, figuring out the requirements, running the procurement, gating Britomart and Newmarket and (eventually) Manukau, installing all those ticket machines and readers and all the other hardware on all the other stations, ferry terminals and busses (and you’ve got to wire them up and securely attach them), paying for hordes of “AT Ambassadors”, paying for the stylish jackets for the AT Ambassadors, training the AT Ambassadors, not to mention training all the drivers, conductors, station staff, call centre staff and so on, paying for lawyers and accountants and graphic designers and photographers and web designers and copywriters and public relations people, running the pilot scheme, mailing all the Hop cards out (someone has to stuff all those envelopes), checking whether people who claim to have an old Hop card really do, checking that people only get a replacement Hop card once, connecting thousands of devices to the Internet, testing all of this stuff to make sure it works, fixing it when it inevitably doesn’t, fixing all the stuff that gets vandalised, doing progress reports and briefings and presentations for the Mayor and councillors and the board of Auckland Transport and your department manager and NZTA and MoT and the Minister of Transport and answering questions from the all of those people, plus the public and the opposition spokesman for Transport.

      And that’s what I could think of in a few minutes, as someone who has nothing to do with the project. I’ve doubtless missed a bunch of huge obvious things and loads more tiny little things that add up. People always think other people’s jobs are really simple, but really, everything in life is unexpectedly complex and hard and expensive and cannot, in fact, be done by Friday.

      http://steve-yegge.blogspot.co.nz/2009/04/have-you-ever-legalized-marijuana.html

      1. If the bus/train/ferry companies where just tasked (as part of there licence to operate) with agreeing on a standard and implementing it, it would have cost far less than $100 mil, and would already be going! There is no need for the above list of “public information” B.S. It’s not a new problem, it has been done before, it just needs to be done

        1. The documents show that the bus companies did get together and offer a solution however it was rejected as it didn’t include the option of integrated fares, something that was known to be essential to building a network like wee are now meant to be getting with the RPTP.

  8. “A better question would be why do we need to spend $100 million on an integrated ticket system for Auckland when an open source solution exits – http://www.calypsonet-asso.org/

    Calypso is not open source, it is a licensed and patented solution, sure the is run by a not for profit foundation, but this is still a proprietory system,

  9. While the machines may cost a grand or so, the first one would have cost god knows how many millions in r and d, associated risk etc.

    If the mark up is as fat as is being suggested why were there not dozens of companies world wide undercutting with a better cheaper produce? Because this rant site is conspiracy and opinion macerating as informed fact (which is why I suspect so many people love to read it)!

    :0)

    1. Readers don’t cost a grand or so, they cost less than $100 (it’s a very simple micro-controller with Wifi and a RFID radio, and not much else). It is the installation that cost that makes it closer to $1000

      No R and D is required, this is old technology, it has been done before many, many times. Why does AT have to do it differently?

      1. not saying it happened here (clearly in some respects it didn’t) but “not invented here” syndrome has a proud heritage in Auckland

    2. Grum, please don’t take a comment from one or two readers and assume it’s everyone who follows the blog! But do take the history of this event with snapper, the cost is fair given the complex project. Take Melbourne for an example of how a project like this could balloon easily. Price from Thales was right, the extra cost due to snapper was unfortunate, but could of been worse.

      1. What are the complex parts of the project. It’s all been done before. the only complex part I know of is navigating the Tender process!

        1. Every city that I know of which has rolled out a completely new system has had teething problems. So expecting AT to get it perfect is a bit harsh. Christchurch only has a system on buses, Adelaide is currently rolling out there’s, but there system doesn’t involve a behavioral change as Auckland’s does.

          In Auckland you have;

          Contractual
          Behavior Change for conductors
          Behavior Change for customers
          Equipment roll out on different companies services – done before but still complex
          Fares (stages) vs. New Fare system – coming later

          Off the top of my head.

        2. Hmmm “Behavior Change”, sounds like marketing speak…. I think people are not that stupid. They have had fare changes before, a decent web site and FB update and some posters should handle that for most. 50% already use a smart card, so no change there. Anyone who has been overseas has probably used train pass systems (kiwis travel a lot). Equipment should be plug into 12V and forget, if it needs more than this, they have got the wrong stuff

        3. Plugging into a 12V system would mean equipment would have to be mobile data equip, slow and expensive! Most people don’t follow HOP or Auckland Transport on Facebook, and how are they supposed to know there is a website without the other marketing? Fare changes before consisted of rising the price of a stage by 50c.

          Obviously people are not as intelligent as yourself there pete…where not all IT geniuses…but in terms of Auckland’s transport it’s a massive change for the users. Prob less than 20% of people used smart cards on public transport in Auckland before so trying to claim it’s not a big change is pretty far fetched.

        4. Pete, you should have tendered for the job yourself, apparently you would have done it better, faster and cheaper. Shame on you for holding back your amazing IT project deployment abilities when they could have been used to save the city plenty of money!

  10. The Christchurch Metrocard isn’t a smartcard system in the way Auckland’s or Wellington’s is. The reason it works so simply is because it is a single zone, single fare, tag on but not off system. It doesn’t calculate fares the way the Snapper or Thales systems do, and if you’re travelling beyond the Christchurch zone (e.g. into the Waimakariri or Selwyn districts) then the driver needs to manipulate the console before you can tag on.

    Technically, it’s not really different from the Howick & Eastern Ezi Pass, so probably best to keep it out of the Snapper/Thales discussion.

  11. It doesn’t really matter what Christchurch, Wellington or anyone else was using as all bidders were were assessed in a competitive tender process designed to get the best solution. Many of the other documents talk in detail about the NZTA review of the tender process and that they found no issue with the estimated costs, how the tender was conducted or the outcome of it. What is interesting is one part indicates that the price Snapper quoted in the media after they lost was not the price they offered in the tender.

    1. Thales and Snapper cards are basically the same. The difference is purely implementation and software, and should be easy to adjust given flexibility and not having one party unnecessarily dictating pointless rules. The float is not big deal, it could theoretically earn $100,000 a year, so what. Most providers including Thales will have visualisation on tag on/off data, what are they going to do with it, who cares? I figured the Government thought it would be common sense to piggy back off an already running system, wouldn’t you?

      1. The blatant favoritism shown to Snapper/Infratil is the issue “Pete”. If Government wanted a truely open standard system that allows for various suppliers, then why did NZTA write to just Snapper immediately after Joyce met with Infratil and Snapper? What about Parkeon or any of the other original AIFS bidders? The next logical question to ask is was Joyce or the National Party paid for this, or was it just mates helping each other out? Can other Government contracts be overturned by a visit to The Beehive? Does NZTA apply any business ethics to its contract tendering processes, or does it simply award contracts to the Minister’s favourites? What about the RoNS? Is that all above board? How the hell would anyone know, given that Gerry Brownlee now has a history of misleading the house?

        That’s the issue Pete. How far has the rot set in?

        1. Have you actually read this post? Thales was chosen by ARTA as the result of an international tender based on a multi-criteria evaluation of a number of vendors. Snapper was one of them. NZTA did a full review of the tender process after Snapper complained. Their conclusion is stated above – “Thales looks to have been a sound choice of preferred tenderer”, and they backed ARTA’s decision. Then Snapper go to Steven Joyce, who instructs NZTA to include Snapper in AIFS despite being excluded by both ARTA and NZTA. So now I’m asking what otr projects has Joyce interfered with? Maybe this was a one off, maybe it wasn’t. But at this point we don’t know. We should.

        2. Thales was chosen by the tender process and was confirmed as the preferred tenderer on two separate independent reviews. You literally cannot get more decisive than that.

  12. Pete, as everyone is trying to tell you, the rot is in the integrity of the tender process when a company like Snapper can lose the tender, lose an appeal over the tender process and yet still worm its way into the project. If anyone should be taking legal action, it should be the commuters of Auckland against Snapper for slowing down the whole integrated ticketing programme after losing the tender FAIR AND SQUARE! Afterall it’s not a big assumption to say that had Snapper not been able to worm its way into the project we would have the AT Hop card on all three modes of PT by now. Basically Snapper’s poking its nose in has probably cost Auckland commuters more than one year of having an intergrated ticketing card at their fingertips. I actually wonder if it would have been better for AT to threaten legal action against Snapper after the appeal against the tender process had been dismissed. At least it would’ve probably got Snapper out of the way for good early on.

    1. Don’t get me wrong, I think most tender processes are corrupt and stupid anyway. Bases on other Thales systems around the world being slow to implement and way over budget, I don’t think we need to worry about Snapper. Snapper is based on a perfectly good system (not great, but functional), why not just extend it. Anyone whom knows anything about RFID and computers would understand that all the systems are more or less the same, it’s just some fancy software that makes the difference, and that could be developed in NZ, hire Xero any other company that has half a clue

      1. Snapper is certainly not great, so why should we extend it? I think we deserve to have a great system. Snapper is very slow and difficult to use and difficult to top-up.

        But regardless of any of that, as others have pointed out, the issue is not which card/system is better, but that Thales was chosen over Snapper twice and then Snapper got into the project anyway due to interference by a minister. That’s where the corruption is, not in the tender process.

        (I sort of think replying to this guy is a waste of time, he doesn’t seem to be listening..)

        1. I am just more worried about another $100 mil being wasted. If you think tender processes are less corrupt than meddling politicians your have missed the point too

        2. Very sorry for being rude, I shouldn’t leave comments late at night.

          I don’t think I have missed the point, as the post was about political interference. It may be that the tender process could be improved, but that’s a separate issue. The fact that there was an appeal which supported the outcome of the tender process suggests that the tender process was looking at the right stuff. The cost does seem high, but I’m not an expert so I don’t know what’s reasonable for a system like this. The tender process obviously looked at the type of system that was preferred and Snapper didn’t meet the requirements. A ‘free’ (I’m sure there would still have been costs associated) system isn’t worth the hassle if it doesn’t work properly.

        3. How can the tender process be improved when we had two separate reviews and a appeal confirm that Thales had the best product/solution?

        4. You got to wonder who Thales bribed? e.g. Thales provided the card system for the Netherlands, quoted as €200 mil, finished at over €1.1+ bil. The Netherlands are only four times bigger (pop. wise) than NZ, we can’t afford to pay $500 mil for a NZ wide system!

        5. Pete – The expected cost when the tender was let was $100m for the system and running costs for 10 years. It compared favourably against the offer that Snapper put up which despite what they said in the media later, wasn’t free as the operators would have paid it and therefore the cost would have come through the system from bus operating subsidies. The tender estimated that the Thales solution would cost $100m and the only change to that has been as a result of the problems that have occurred with Snapper.

        6. So far the only cost blow out has been caused by the snapper fiasco, so it’s actually been quite successful from Thales end so far. So can’t see the problem here as of yet, and the main failure point of their system you have indicated has been passed. The back-end part is complete, rail is online and just the buses to go.

  13. I think NZ Bus is interested in retaining their system because:

    (1) they have made a considerable investment in Snapper with additional investors,
    (2) there is the marketing and revenue generation potential with retailers & other public transport operators (similar to how debit & credit cards work – on a commission basis), and
    (3) the level of control that NZ Bus has over the data.

    Whether NZTA or AT can contain the costs of implementing Thales’s smart card system depends on

    (1) the level of customization required,
    (2) how the project is managed, and
    (3) how strict they would hold their contractor responsible during the course of the implementation.

    In sum, the more customizations and the more decision changes being made during implementation will incur more costs.

  14. NZ Politics runs on cronyism, why is anybody surprised,Members of Parliament have a reputation below that of used car salespeople.
    You should all be surprised when the peasantry and citizens of NZ do not get sold a lemon by them.

Leave a Reply

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