• NAR OIA Request to NZ Police (pdf, ~550KB)

(New Zealand Police Header)

10 June 2008

Gavin Treadgold

Dear Mr Treadgold

Thank you for your Official Information Act request of 12 May 2008 about the National Address Register (NAR) project.

You have asked for a range of tendered prices and would like to know, so as to get a better understanding of the economics, what the lowest and highest price tenders were for the project. E.g. the range of prices that were tendered. You state that you believe that this is not commercially sensitive information as no tender has been awarded, and no vendors are being associated with the tendered price range.

I can advise that the prices received ranged between $9M and $48M for conforming tender responses.

You have asked for the number of tenders and I can advise that there were four responses to the RFP. One tender was evaluated as being non compliant to a mandatory requirement.

Yours sincerely

Jill Barclay
Technology Manager: GIS
Information, Communications and Technology Service Centre

It was interesting watching the request being passed from agency to agency (although it was understandable). I submitted the request to Police as they were leading this stage of the NAR. The request was passed then to the State Services Commission, and finally on to LINZ, who released the document quite quickly.

Of interest in this document will be comparing government’s estimate of the cost of implementing the NAR over a five year period – versus the submitted tenders. I, of course, am trying to find out the price range of the tenders for comparison.

It is still really disappointing that this project has failed, it had so much potential to deliver a fantastic  authoritative dataset to New Zealanders.

• National Address Register – Moving Forward August 2007 (pdf, ~200KB)

This project had the intention of providing a single national authoritative dataset for roads, addresses and placename information. The potential of this project was to deliver a free dataset that all organisations and individuals in New Zealand were free to use. This would have made a fantastic resource, and had the potential to consolidate a number of mapping projects, and could have greatly simplified the work associated with project such as the NZ Open GPS Maps project, as the NAR would have provided a single national focal point for feedback and correction of road and address information.

The cynic in me says that the reason this project failed was because of the commercial interests in existing roading datasets. Currently there are multiple roading datasets from different providers, and they are making very good money from these. Some roading datasets sell for six figure sums on an annual basis. Naturally, very few organisations can afford these prices, so only large Government agencies tend to be able to purchase them. Suffice to say, these datasets are different, and there is not a single authoritative dataset amongst them.

The NAR had the potential to create a single, free and authoritative road, address and placename dataset. Tenders were invited for the project, and there was going to be only one organisation to win the tender. As a result, all but one of the current commercial providers stood to lose their revenue streams from their roading datasets. As you will see in the notification below, the tenders were too expensive. I believe that it was in all the commercial vendors interests to put in high tender prices to ensure that the NAR did not go ahead, and that they could protect their existing revenue streams rather than risk missing the tender and losing it all.

The upshot of this is that my faith in the Government to provide geospatial information to its citizens is now close to zero. If they are not capable of producing a single authoritative roading dataset (arguably one of the most important sets of spatial information as it defines most of our physical connectivity) then there is little hope of them being able to deliver any useful spatial information to citizens.

As the NZ Open GPS Maps, Zenbu and NZ Open Street Map projects have shown us, a volunteer community can develop products faster and cheaper than commercial or government organisations, and over time they will have better quality as well.

I believe the time has come for us to build more volunteer communities to provide spatial information that our Government is failing to provide to us. No longer can we wait upon them, rather we must build it ourselves. There are four key areas that we need to focus on.

1. Raw data collection – taking our GPS units out into the real world and collecting and sharing data. Collecting track logs and uploading these to the OpenStreetMap or NZ Open GPS Maps projects. Providing waypoints to OSM and Zenbu. Please – if you haven’t already, consider donating some time and information to these projects so that they have raw data to work with. This ‘field survey’ work is essential to creating our own spatial information resources. (I would particularly encourage geocachers to contribute their tracklogs if at all possible as we tend to travel a little more than others)

2. Mapping – converting the data collected in the field to information. Creating vectors for road lines, adding street name, directions, speeds. Using your local knowledge to map the community around you.

3. End products – converting the spatial information into a form suitable for others to use, for example the NZ Open GPS Maps project producing Garmin map files that can be loaded into GPS units.

4. Distribution – due to the large quantities of information involved, we may need to look at creating an ad-hoc network of individuals and websites to share the vast quantities of information about our country via torrents or similar P2P mechanism.

I believe the time has come for all those that want better access to spatial information to go out there and be a part of collecting, and building it. We can’t wait for Government to build it for us, so we will have to do it ourselves.

Let’s get started.

SUBJECT: NAR PROJECT TERMINATED

The National Address Register (NAR) project is a cross-government initiative set up to develop infrastructure to improve the provision of address, road and place name information for government agencies, businesses and the wider community.

The project is over-seen by a Steering Committee comprised of representatives of key stakeholders from central government, local government and emergency services agencies.

An integral part of developing a business model and business case for the National Address Register (NAR) was to assess whether there was a supplier able to provide the relevant services and to identify the likely costs. An RFP process was chosen as the most effective way of identifying both of these.

Following assessment of the tender proposals, the Steering Committee has decided to terminate the project. Despite the project showing considerable potential to reduce duplication across government and reduce costs, it is too expensive to proceed with in its current form.

Further investigation into the need for, and the most cost-effective way of providing address, road and place information, will be led by the New Zealand Geospatial Office, within their mandate under the NZ Geospatial Strategy. This work will include determining the optimal role for the Crown, local government and the private sector. Brendon Whiteman, Director New Zealand Geospatial Office - brendon.whiteman@linz.govt.nz; will be happy to answer any queries that you may have in relation to these activities in the context of the overall work programme of the Geospatial Office.

It is expected that agencies will continue with existing arrangements they have for the purchase of this location data, from the commercial sector.

Nancy McBeth of the State Services Commission, is preparing a Lessons Learnt report on the NAR project. If you have some views that you would like considered in that report, please contact her at nancy.mcbeth@ssc.govt.nz by 31 May 2008.

On his return from Annual Leave next week, Laurence Millar, Chair of the NAR Steering Committee will formally write to your Chief Executive to advise of the decision.

Regards
Jill Barclay
Operational Owner NAR project
NZ Police

Received from a public email list I subscribe to.

There are broad level maps available, for example Rangitata, as well as a web mapping solution – although as I write this, it doesn’t appear to be displaying the proposed boundaries to any significant level of detail.

Disappointed at the options of trying to compare the current and proposed electoral boundaries, I sent some feedback to the website asking if they would consider making the information available in a more suitable format.

Hi,

I have just visited this page on the proposed electoral boundaries.

Is there any chance that you could provide the boundaries in
geospatial formats so that the people can view it in say Google
Earth? Publishing shapefiles for the proposed changes would also be
useful for those people that are GIS-savvy like myself and would like
to perform a bit more analysis.

Thanks in advance for the consideration.

A few days later on the 8th of May, I received this reply from the Electoral Commission.

Dear Mr Treadgold

The Commission is required under the Electoral Act to make the proposed boundaries available in a form that ensures they are defined accurately. This is to facilitate public inspection of the proposals and the making of objections and counter objections to them. The Commission is not able to provide the boundaries in geospatial format because it would not be able to control the continued accuracy of the boundaries released in this format.

Yours sincerely

name removed
For the Secretary of the Representation Commission

I replied with the following.

Thank you for the reply name removed.

There is a need to strike a balance between making the information
available, and making it available in a form that can be used for
analysis. Currently, there is no demographic information provided on
the interactive mapping of the Elections website. If there are no
plans to make this available, then some means is required to make the
proposed boundaries available in a form that opens them up to more
rigourous analysis, so that more thorough consultation over proposed
boundary changes can occur.

Can I make a suggestion to the Representation Commission that they
should consider making the proposed boundaries available as a web
service as suggested by the e-GIF Geospatial Information Standards
that are part of the NZ E-Government Programme? It should also
include the use of metadata to define the accuracy and other relevant
metadata. Land Information NZ could naturally provide more
information here.

e-GIF Geospatial Information Standards

A suitable compromise may be able to be reached by providing the
information via a Web Feature Service (WFS) that allows control and
accuracy to be maintained, but at the same time allows GIS users to
combine the trusted information directly from your server, with
demographic and other relevant information (that is currently not
being provided on the Elections website) in their GIS to allow proper
analysis to occur. This mechanism works by allowing GIS software to
connect directly to a server (under Elections control) that allows
downloading of the boundaries directly into the GIS software so that
analysis can take place with other data layers.

If the boundaries are not provided in a geospatial format, I would
suggest that it will not be long before individuals sit down and
digitise the boundaries themselves to meet the need. This process
would likely produce more errors, errors that may not have occurred
if the data was provided in suitable formats in the first place by
the responsible agency.

This would most likely occur in the lead up to the next election as
interest in the electorates increases. I am aware of individuals that
have downloaded the whole topographic database of New Zealand from
LINZ’s TopoOnline website just so that the data can be used to
produce maps for use in mapping GPS units because the current
offering in the market at the time were abysmal (I am currently
President of the NZ Recreational GPS Society ).

It would make more sense for the Commission to consider the
publication of authoritative boundaries in a manner that pre-empts
moves to produce boundaries by alternative*, labour-intensive and
error-prone means.

Kind regards

Gavin Treadgold

* There are a number of countries that don’t provide road information
in a geospatial format – such as Australia[1] and the UK. In the past
few years, significant data capturing efforts[2] have been undertaken
in numerous countries to create duplicate datasets because the
Government is not meeting the demand.

[1] http://www.gpsoz.com.au/tracks4australia/
[2] http://www.openstreetmap.org/

I believe this is a very important issue, perhaps one of the most important spatial issues that we could face – that of making the data available in a means that allows detailed analysis of any proposed electoral changes to occur. If you feel strongly about this, then I would recommend that you contact the Electoral Commission and ask them to make the information available.

I’ll keep you informed of replies.