Documents say the country's speed cameras were already a few years old when police put them in - and are now so ineffective they miss catching a lot of drivers.
The old mobile and fixed cameras are a bit more than half as effective as new ones, and the red-light cameras 15 times less effective.
Most of the red-light cameras are owned by Auckland Transport, not police.
Though they were installed between nine and seven years ago, police actually bought them well before that, and they "were likely developed a further 10 - 20 years before that".
This was revealed in a summary of a confidential due diligence report by accounting firm PWC to the Transport Agency late last year.
Only approximately 70 percent of images [are] being captured accurately and 30 percent of non-compliant drivers [are] not being processed
The worst performing cameras were those used against red-light runners; they snap just one offender where a new one would snap 15.
"A large proportion of police safety cameras uses older generation technology, which is less efficient, is less effective, and generates sub-optimal road user compliance (that is, doesn't deter speeding)," said the summary in a Waka Kotahi business case that paved the way for a significant expansion of safety cameras from now till 2030.
RNZ had asked police why they installed old cameras in the first place, given the less effective a camera network was, the fewer lives it saved, according to NZTA reports.
Waka Kotahi was taking over a network that was on its last legs, which included the 100 or so old police cameras.
The support system behind the cameras was so dated it required manual downloading of images.
In addition, the infringement processing system was also old, slow and must be replaced entirely, reports said.
The agency was testing new cameras in Auckland.
PWC did due diligence on the old cameras and on the police's infringement processing system, called PIPS.
"While police installed new cameras over 2014-2018, most used first-generation technology that can catch fewer non-compliant drivers," the business case said of PWC's findings.
"These older cameras capture fewer incidents and allow more non-compliant drivers to get away with speeding on the roads, leading to a higher level of DSIs [deaths and serious injuries] than would be the case with new generation technology."
Cameras were installed years after they were bought, and longer still since the tech had been developed, it said.
"For example, static cameras installed by police eight years ago were procured a few years before being implemented and were likely developed a further 10-20 years before that," the report said.
Over a six-year period to September 2021:
- Each of the country's 54 fixed cameras detected 7600 incidents, when new cameras would have detected 15,000 (96 percent more)
- For 45 red-light cameras, 196 versus 3100 (1482 percent more)
- For 43 mobile cameras, 8500 versus 15,000 (75 percent more)
Police told RNZ they bought 48 fixed cameras from 2014 to 2018 "and quickly went about installing them".
"At the time of procurement, these cameras were the latest technology available," director of the national road policing centre, Superintendent Steve Greally, said in a statement.
"These cameras have fulfilled their purpose to high standard over that time.
"However, technology has improved significantly during this period."
RNZ has asked police for more information to back up these assertions, in light of the due diligence findings reported to NZTA.
PWC's findings were that the three camera types had four-to-six years of life left in them.
Auckland Transport owns 42 of the redlight cameras, police just three.
The red-light cameras could also detect speeders - except this function was not enabled, the report said.
This was because the red-light cameras could not generate an infringement notice for speeding, Greally said.
If they had been able to, deaths could be cut by up to a quarter at intersections, the NZTA report said.
The old fixed cameras had very limited ability to recognise number plates - though this ANPR technology was now widespread on privately-owned CCTV cameras at major shops and service stations, which often fed images to the police including of shoplifters' cars.
The old mobile cameras could not tell if vehicles were crossing between lanes, "which limits prosecution".
The network had no cameras that could measure speed between two points, and ping a driver if the average was over the limit.
Overseas these were "highly effective", PWC said.
A law change was needed here to use these "average speed" cameras; the other barrier - high cost - had been overcome, with per kilometre costs falling from $1.5 million to less than $100,000 in the last few years.
The documents showed Waka Kotahi testing or planning to use cameras from two Australian suppliers.
Calibrating the point-to-point cameras would be key - and this function looked likely to pass from police to the companies.
The transfer of the cameras would cost some jobs from among the 150 or so police staff who worked on the camera or infringement systems, or some may transfer to NZTA. Unions were involved in talks, the business case said.
PWC found both police systems - for cameras and processing infringements - were "at end of life".
This caused delays in issuing of tickets, which eroded deterrence as this depended in part on a swift penalty, the report said.
The tech is one thing, strategy another, and that has been lacking too, the report said.
"Safety cameras have not been systematically targeted at high-risk parts of the road network to reduce DSIs," it said.
The last formal review of the police's safety camera site strategy dates back to 2002.
Then, the Auditor-General found: "Police had no national standard or policy on the use of deployment plans for determining speed camera deployment on a day-to-day basis".
Camera deployment was left up to district commanders or operators, who often chose mobile camera sites "for convenience".