the Irish GI Community. Richard Cantwell Senior GIS Consultant / Technical Manager email@example.com www.gamma.ie Premier Partner Premier Partner Good Afternoon. I’m Richard Cantwell from GAMMA. I’ve been working in the GIS field here in Ireland and abroad for 20 years, as have GAMMA GAMMA are Ireland’s leading Geocoding and GeoDemographics consultancy and we are the Premier Partner for Pitney Bowes MapInfo for Ireland. I’m here today to talk about Postcodes. Something that is very important to all of us who work with GI in Ireland, and something that has had a fraught history, but we are beginning to see some light at the end of the tunnel. 1
even in the same building, to write their addresses differently: 56 Woodbrook Sq., Castleknock, Dublin 15 42 Woodbrook Sq., Diswellstown Rd., Clonsilla, Dublin 15 34 Woodbrook Sq., Carpenterstown, Dublin 15 So, Why Postcodes? As I’m sure you all know, Ireland is the only EU and OECD country without a postal geography. As such our addressing has been ad-hoc. At GAMMA we have geocoded many millions of address records and have come across huge numbers of issues like this, where people who live in the same building record their address very differently. 3
by any means, as I’m sure most of you know. People use townlands as the basis of their address, meaning huge numbers of people have the same address – dots with the same colour in this image share the same address. This is a big problem, not only for delivery of post, but for service planning, customer analysis, and so on. 6
big problem – how do you avoid duplication? Different 'households' may live at the same address – as we've just seen. The same household can have different addresses – depending on who entered the address De duping this, an essential requirement for a range of analyses, is problematic in the extreme. 7
4 Million P.A. One example: The annual cost to the HSE of Missed Appointments is estimated at over Eur 30m. One study conducted in the Mid West Region pointed to Address Issues causing in the order of 13% of this problem, an annual cost of Eur4m 8
been complex and drawn out. I'm not going to detail every event in the recent history of the Government's postcode project, but, broadly speaking, this is what has been happening, and what is next.. 11
with the establishment of a ‘Postcode Working Group’ which commissionned a report which recommended their introduction. The then Minister, Noel Dempsey, announced they would be introduced on 1st Jan 2008. 12
Announced In Sept 2009 Minister Ryan announced that the system would be implemented by the beginning of 2011, some examples, of area based codes, were issued in the accomanying press release. This was followed by the issue of Tender Documents to select a consultant to advise the department in Feb 2010 14
Announced Oireachtas Report on Postcodes Tender Cancelled The tender issued in Feb was subsequently cancelled, due to ‘an unspecified technical error’. The tender was subsequently re-issued and awarded to PA Consulting in September 2010. 16
Announced Oireachtas Report on Postcodes Tender Cancelled New Tender Announced In November 2010 the 2nd phase began, with a tender announced to select the PLMH – Postcode Licence Management Holder, who will be responsible for establishing, operating and maintaining the Postcode system. 17
Announced Oireachtas Report on Postcodes Tender Cancelled New Tender Announced Pre-Qualification Questionnaire Published Pre-Qualified Tenderers Selected 3 consortiums were selected from the pool of applicants following the pre-qual. 19
Announced Oireachtas Report on Postcodes Tender Cancelled New Tender Announced Pre-Qualification Questionnaire Published Pre-Qualified Tenderers Selected ‘Competitive Dialogue’ / Specification & Design And these consortia have been engaged in 'competitive dialogue' with the Department of Communications since. As far as I am aware this 'dialogue' includes the specification and design of the postcode. We should see this 'shortly' Indeed, when I proposed this talk I was hoping it would have been published, but, alas... 20
'soon', and by soon I mean a matter of weeks (hopefully) But delays are part of the story at this point, so it could be months. Once the spec is published the pre-qualified tenderers will then formulate their responses.
‘go-ahead’ Of course the slection / announcement of a winner will be subject to a 'go-ahead' from the cabinet – as we have been repeatedly told by the current and previous ministers. This is by no means certain, especially in the current climate. We, as GI professionals, can clearly see the advantages, but they're not so easily seen by the voting public.
‘go-ahead’ Roll out The 'roll out' can be estimated as taking 18-24 months, minimum. There will be a lot of retro-fitting, as well as the usual 'stationery' changes So, printing companies will be happy, as will that old guy who knows Cobol and was last seen charging the banks an arm and a leg to prep them for the millennium bug. He probably has a Ferrari branded zimmerframe at this point.
...is [it] part of the specification that it will be able to deliver to every household and house in the State, no matter which boreen or apartment complex one lives in...? Deputy Pat Rabbitte: That is the specification because anything else would be unconscionable. It shows how complex and demanding it will be because the notion of some sections of the country, merely because they may be more remote or isolated areas, not being beneficiaries of the system could not be countenanced. www.kildarestreet.com/sendebates/?id=2012-05-15.270.0 In May, for example, the clearest indication yet that the postcode will be at Building level, at least.
56 Woodbrook Sq., Castleknock, Dublin 15 D15 4AQ3F 42 Woodbrook Sq., Diswellstown Rd., Clonsilla, Dublin 15 D15 4AQ31 34 Woodbrook Sq., Carpenterstown, Dublin 15 D15 4AQ4H Or at address point level. Again, in Urban Areas with house numbers and street names this is pretty much a solved problem already – using GeoDirectory and AutoAddress.
understanding is that the PLMH will produce a 'PAD' which will not be free. System needs to be maintained and this needs to be charged for. Will be free to the 'end user' of course – people won't have to pay to find their postcode.
you might not be able to specifiy the format for a few months, at least. If possible, hold off until the spec is announced. If your timeframe is longer, then you need to do nothing right now. But getting the number for the zimmerframe wielding COBOL guy might be a good idea.
we can expect postcodes to make a big difference, especially in rural areas. The oft quoted issues of ‘findability’ will be solved by a building level postcode. After all, 32% of Irish Addresses are not unique, which is something that absolutely astonishes international clients of ours.
the forgotten stepchild of the organisation – they’re messy, they’re hard to manage and they’re often decentralised into Silos, with different parts of the organisation having their own databases with their own requirements tacked on. Therein lies a massive opportunity.
A centralised, managed and deduplicated address database unlocks a huge amount of intelligence. There is a big opportunity for GI managers in organisations to enchance their visibility and their service offering. Just think if a local authority, for example, had a centralised system with every connection with their citizens logged and recorded, how much duplication of effort and ‘lost / missing’ records would that save? How about the HSE and their missed appointments issue?
to solve all of these issues on their own. They will be very useful addition to the data landscape, and the toolkit for most of us who work with spatial data. They will make many of these things significantly easier, or even possible. But for a significant proportion of use cases the solutions currently available, such as AutoAddress and GeoDirectory may already be sufficient.