Released
Proactively
Background and where we are now.
2. In 2021/2022 NZTA and Police tested two types of plate with silver or white figures1 on a retroreflective black
background2. NZTA approved them and they have been on sale through KiwiPlates3 (the sole approved
personalised plate marketer) since September 2022. Overs 9(2)(b)(ii) of them have been issued. In this memo
they are called “the problem plates”.
3. NZTA has considered revoking the approval of the problem plates previously4 with the decision to maintain
current approvals maintained. This memo should be read with the context of the previous Internal link to d4 as
referenced.
4. This testing involved driving vehicles fitted with the problem plates past various cameras. The cameras
included Police’s current speed cameras (Redflex NK7s) the new model of camera the safety camera
programme will use (Redflex Halos), and the cameras used in weigh-in-motion systems and on toll road
gantries.
5. After the approvals were made, Police, NZTA staff, and the public, reported many cases where the figures on
the problem plates could be hard to read or become essentially invisible to the naked eye under some quite
common conditions. Police have repeatedly requested we withdraw approval for the problem plates.
6. We have scientifically and measurably characterised the readability / invisibility problem. The retroreflectivity
expert we engaged says that the problem lies with the retroreflective background’s performance relative to the
non-retroreflective white and silver materials used to form the figures in the problem plates.
7. Under certain circumstances the retroreflective feature of the background material can become partially
initiated and in doing so it develops a silver-white colour that closely matches the non-retroreflective colour of
the figures. The colour contrast between the two products reduces to near zero making figures hard to read or
indistinguishable from the background.
Released
8. It’s no surprise that this was not picked up in drive-by testing. The cameras used in the drive-by tests were
those specifically designed to pick out number plates for direct use in law enforcement (Police speed cameras,
our new safety cameras, tolling, and weigh-in-motion cameras).
9. The cameras used in those roles have optics specifically designed for reading plates, have their own IR or
visible light illumination, and use configurable and very capable proprietary software. They are designed and
configured so as not to cause problems for police forces or road controlling authorities in markets across the
world. New camera systems can cost around $250,000. They are installed using fine tolerances around angle
of view and flash illumination.
10. Driving by such cameras doesn’t tel you anything about how plates can be seen by the human eye or by less
sophisticated camera set-ups.
11. More generally drive-by tests are a poor way of assessing whether a plate is readable. They are not
quantitative, replicable, or adequately defined. They produce non-quantifiable results under poorly controlled
conditions and should not be used as the primary test for approval.
What we need to do.
Proactively
1 Retroreflectivity is a very different thing from simple reflection and the differences are highly relevant see the short primer on retroreflectivity in
appendix 1
2 The plate with silver figures uses black retroreflective sheeting—3M 680-85 Black and Kurz Silver Hot Stamp foil M72462. The plate with white
figures uses the same material for the background (3M 680-85 Black) and Kurz White Hot Stamp foil B721332.. It is important to understand and
specify the materials used in plates such as these foils. They have characteristics (usually specified in data sheets) and are typically tested against
standards for licence plates retroreflective and materials like ISO 7951 AS/NZS1906.1:2017 and agency adaptations of these like our NZTA
M25:2021 Key steps in applications and approvals for plates should require the materials to be precisely specified . Some of the confusion and mis-
steps to date are generated by failures to specify and understand the behaviour of these materials.
3 KiwiPlates is the trading name of the only entity to have a (non- exclusive) contract and approval to market personalised plates, Multi Market
Services Limited, the ultimate beneficial owner of that entity is Publicis Groupe Holdings B.V. registered in the Netherlands. The KiwiPlates website
shows how they market personalised plates including the problem ones see https://www.kiwiplates.nz/
4 See memo Internal link to document provided in Attachment 2
PAGE
2 OF
11
12. We need to develop a physics and standards-based test procedure for approving the materials used in
registration plates. The test needs to assess whether there is sufficient contrast between figure and base
material. That’s the key failing that led to the current situation, and it’s not covered in existing standards.
13. There is a more immediate need to withdraw approval for the problem plates and to approve a new form of
black plate that does not have visibility issues.
14. s 9(2)(b)(ii)
. It would also significantly reduce the revenue share paid to NZTA, which is
the only funding source for the Community Road Safety Fund5.
15. s 9(2)(h)
16. s 9(2)(h)
17. s 9(2)(j)
18. The decision can’t take effect immediately. It will take the plate manufacturer LicenSys until 30 November
2024 to install the equipment and make the necessary software changes to produce the replacement plates.
19. This memo recommends a solution that provides certainty for KiwiPlates and recognises the lead times
needed for the plate manufacturer:
Released
1. Make the decisions to revoke approval for the problem plates and approve the replacement plates
now.
2. Establish that these decisions will come into effect on 30 November 2024.
20. The replacement plates would have a non-retroreflective black background and white retroreflective figures 7.
The material for the figures is the same white retro reflective material used in ordinary plates. Put another way
the replacement plates use known materials and are essentially the inverse of the ordinary plates found on
most vehicles.
21. Interim test
22. The readability issues with the problem plates are, of themselves, a basis for withdrawing approval.
Registration plates are (with licensing labels) one of the physical tokens that serve to alert people to the
registration status of a vehicle. They are critical to the enforcement of traffic offences and to investigating other
criminal offending. They are also critical in identifying who may be liable for civil action such as after accidents.
The plates are simply not doing the job required of them if they are difficult to read or the figures on them are
essentially invisible.
23. The work to develop a physics and standards-based test has already produced results that allow us to
prescribe an interim test on which a decision to withdraw approval can be made. The interim test requires:
1. Using a retroreflective material that conforms with relevant standards.
2. Having a contrast ratio between figure and base material in an appropriate range (>5:1 and < 11:1)
Proactively
24. The problem plates fail this test while the replacement plates pass it (see appendix 3 for details).
5 See https://www.nzta.govt.nz/safety/what-waka-kotahi-is-doing/community-road-safety-fund/ for an outline of how that fund is used
6 See memo Internal link to document provided in Attachment 2
7 the materials are figures in a white retroreflective material (being 3M 4750 or 3M 4780) and a non-retroreflective black background (opaque
digital Ink) this is the inverse of the currently issued ordinary black on white plates
PAGE
3 OF
11
What parties think of this approach
25. KiwiPlates (the personalised plate marketer) and Licensys8 (the plate manufacturer) support withdrawing
approval of the problem plates and simultaneously approving the replacement plates as from 30 November
2024.
26. Internally Scott Henshilwood, Manager, Regulatory Contracts and Kim Gaffaney, Contracts Officer with hands-
on responsibility for KiwiPlates and LicenSys and Eric van der Plank, Manager Commercial Services are in
favour.
27. Police will support the withdrawal of the problem plates (and have repeatedly sought this) but the officer in
charge of CCTV and ANPR (Automated Number Plate Recognition) technology development at the Police
Visual Information Centre at Manukau has some has significant concerns about the replacement plates’
readability.
28. He conducted two series of tests one of which appear, to show significant readability problems with the ANPR
cameras routinely accessed by Police. That report was seen in earlier NZTA work as showing the replacement
plates “test even worse than the problem plates”.
29. In fact, the issue is as much about the capability of the cameras as it is about the plates. Part of the planned
future work is to engage with ANPR camera operators and suppliers to allow them to adapt as far as possible
to the replacement plates.
The one difficult case: “Police ANPR cameras”
30. It is critical to understand what the “Police ANPR cameras” are, and why we can’t possibly devise a system for
plate approval that requires readability by all or most of the cameras involved. The term “Police ANPR
cameras” is very misleading since few of the cameras belong to Police. Rather they mostly belong to two
network operators Safer Cities (which has cameras operated by local authorities and local business
Released
associations) and Auror (which has its cameras mostly at major fuel retailer’s sites). These vary greatly in
quality in terms of optics and ANPR software. There is an explainer on this in appendix 2
.
31. The most appropriate policy response for that industry is to set standards for plates and make those standards
public. We also need to show ANPR camera suppliers and operators examples of any new plates (and their
specs) with sufficient time to adapt their systems.
32. We will need to be clear with Police (and internally) that:
1. the way forward is objective physics and standards-based testing, not solely drive-by tests.
2. we are engaging with ANPR operators and manufacturers on the readability of the replacement
plates.
3. these processes will facilitate but do not guarantee readability by all the cameras, firmware, and
software combinations in the ANPR cameras accessed by Police.
4. ultimately, it is an issue for camera operators and suppliers to ensure cameras, over time, are able to
read plates in use.
33. I expect that the work mentioned in para 32 above will allow ANPR operators to make changes to software,
hardware, or systems that can better read the replacement plates. This should, over time, assuage Police
concerns about readability but not eliminate them entirely. We should be aware that some residual risk will
remain.
Proactively
8 LicenSys is Licensys Holdings Limited, the only major plate-maker in New Zealand it (along with an equivalent Australian entity) are ultimately
owned by Erich Utsch AG a German registered company that supplies plate making equipment. LicenSys has the only contract to supply ordinary
registration plates to NZTA agents as well as the personalised plates marketed by KiwiPlates.
PAGE
4 OF
11
Released
Proactively
These processes are about showing camera operators what’s coming. They would not
form part of the approval of the replacement plates or any future plates. Plate approvals
would be solely based on the physics and standards-based test.
3) Incorporate the test described in 1) in NZTA documents and by extension the
arrangements with LicenSys and KiwiPlates via some or all the following:
•
a revised NZTA plate specification
•
an application process for approval of plates specifying the exact materials
they will be made from and their specifications.
•
requiring independent evidence that proposed plates comply with the test.
•
devising criteria for being an independent tester
•
adjustments to other documented procedures
•
prescribing a standard process for alerting camera operators to the approval
of any new plate providing for them to see how it works with their cameras
above recording the results.
•
being very explicit that such demonstrations of new plates are not a part of
the approval process.
•
remove redundant references to MoT and to other documents and
processes.
•
make the documents internally coherent and explicate their relationship to
other documents so there is a single source of truth.
Note that both the contracts with KiwiPlates and LicenSys require them (in slightly different
terms) to comply with NZTA specifications for plates.
4) Reconfigure how NZTA does plate approvals including by —
•
recognising plate approval is a regulatory decision and locating the power to
Released
make it (and related work) in the core of the regulatory group.
•
separating decisions on plate approvals from day-to-day contract and
relationship management
•
Separating assessments against the standards and decisions on plate
approvals from the standard setter.
5) Reconfigure how we appoint the Registrar of Motor Vehicles as a statutory officer
under part 17 LTA 1998.
We should give effect to the intent of the underlying provision by appointing someone
whose span of control at least roughly corresponds to administration and QA for the
register and closely connected issues such as registration plates.
The idea here is that the registrar should have a clear set of responsibilities around
the register and be (to some extent) immune from other (including commercial and
income-based concerns). This would create a useful tension between conflicting
objectives.
Proactively
PAGE
6 OF
11
Approval
Decisions sought
Approval
revoke approval of the problem plates described in detail and using the interim test in appendix 3
YES
with effect from 30 November 2024.
approve under r 34(4) (b) Land Transport (Motor Vehicle Registration and Licensing) Regulations
YES
2011 the replacement plates described in detail and using the interim test in appendix 3 with effect
from 30 November 2024.
Agree to the actions, as set out in table 1, to establish a new personalised plate approval process.
YES
s 9(2)(a)
Signed:
Date: 15/08/2024
Chris Rodley National Manager Regulatory System Design
Note: I make this decision using the delegation to keep this process moving forward, but as captured above, there
needs to be separation between the standard setter function and the assessment and decision making against that
standard. In general, the standard setter should retain the delegation though, but only to ‘set the standard required’
and act as a decision backstop should the regulator and regulated parties disagree on a way forward – essentially
providing for semi-independent decision making.
Released
Proactively
PAGE
7 OF
11
Released
Proactively
Standards for retroreflective materials specify quite precisely the cone of retroreflected light that’s needed and its
colour and intensity.
Retroreflectivity can lead to some strange effects. For instance, the retroreflected light can be of a quite different
colour to the surface as seen in say daylight. That effect can sometimes be seen in road signs.
In this context the black retroreflective sheeting (3M 680-85 Black) in the problem plates can under certain
illumination conditions (very counterintuitively) actually be white. That is at the root of the issues with the problem
plates.
NZTA Requirements for retroreflective materials
We have some quite well-developed standards and practices around retroreflective materials for road signs and
dozens of materials have been approved in that context. Our requirements in relation to plates are less developed.
The NZTA M25:2021 Specification for retroreflective sheeting9 requires that all retroreflective sheeting used in the
manufacture of registration plates must be compliant with AS 1906.1:2017 Standard for retroreflective sheeting10
and be listed in NZTA M25 Listing of approved vehicle registration plate sheeting11. This requires the specified
sheeting properties to be tested both in a materials testing laboratory and at an outdoor weathering testing location
in Northern Queensland to prove performance durability.
The two sheeting colours that have passed these tests are white (3M 4750 and 3M 4780) and yellow ( 3M 4761
lemon yellow 61 and 4780 yellow 81) . The approved white is used for the background on all ordinary registration
plates other than dealer plates which use one or other of the approved yellows.
None of the materials in the problem plates comply with the requirements listed above. There is also an issue about
Released
the standards themselves because they don’t deal explicitly with contrast between figure and ground (the root
problem— this is why we need to add a contrast test to our internal standards.
Appendix 2
“Police ANPR Cameras”
Tom Zotlof NZTA ANPR expert is to review
and if available in time material can be added to elaborate as needed
in particular on the adaptability of these cameras
ANPR involves something that captures images (often in, but not limited to visible wavelengths) and very rapidly
converts that data into formats (there are standards for this) that can be numerical y computed so that the character
sets on plates can be derived.
You can run a quite effective ANPR with a modern Android smartphone and readily available open-source or
proprietary software. A mildly computer literate person can (because standards) add ANPR functionality to many
modern dash cameras. There are active online communities where folk participate in this and there are many
systems for many platforms.
ANPR can be bolted on to many other kinds of cameras such as those used for surveillance by individuals, shops,
Proactively
residents groups, local authorities (often as part of wider safety initiatives) supermarkets, petrol stations, road
controlling authorities, hospitals, airports, and other entities like parking providers.
9 https://www.nzta.govt.nz/assets/resources/retroflective-sheeting/NZTA-M25-2021-Listing-of-approved-vehicle-registration-plate-sheetings.pdf
10 AS NZS 1906.1.2017 Retroreflective sheetingmarked for black plates standards.pdf
11 https://www.nzta.govt.nz/assets/resources/retroflective-sheeting/NZTA-M25-2021-Specification-for-retroreflective-sheeting-and-notes.pdf
PAGE
9 OF
11
Released
Proactively
Out of Scope
Released
Proactively
Out of Scope
PAGE
11 OF
11
