Safe Speed Forums

Reported in http://news.bbc.co.uk/1/hi/england/wear/4723530.stm Wearside police have had a two day ANPR blitz... Laudible in some ways, but when you look at the results, it doesn't say a lot for the accuracy of the databases that they seem to be using...

Extracted from the BBC report:-

Five police vehicles fitted with Automated Number Plate Recognition (ANPR) equipment were parked in Sunderland on Wednesday and Thursday.

...During the operation there were 891 vehicles which were identified on the system as being of interest to police.

Some related to DVLA or insurance offences, but 374 were flagged up as being within the target area of the operation, and were stopped and challenged.

Now that's 891 vehicles "of interest" of which 374 were stopped because they were in the "target" group. As a result of all this activity they made 14 arrests...!

Chief Superintendent Dave Pryer, Sunderland Area Commander, said: "This system allows us to target vehicles which intelligence tells us have been involved in criminal activity.

"This means we are not stopping vehicles belonging to law abiding members of the public but are using our resources efficiently."

I'm afraid that I beg to differ... 14 out of 891 is a "hit rate" of 1.57% - or to put it another way, they were wrong 98.43% of the time. Even being generous and limiting it to the number of vehicles actually stopped, they still only had an affective hit rate of 3.74%, so 96.26% of the cars they stopped contained "law abiding members of the public".

Now, extrapolate this to the proposed nation-wide ANPR system...!

_________________
"Politicians are the same the world over... We build bridges where there aren't any rivers." - Nikita Kruschev

I agree the database and OCR system aren't exactly flawless (not to mention the growing problem of clones plates), but not all road traffic offences are arrestable (e.g. outstanding CC charge/parking fines, no insurance/tax).

That's true... But, according to the BBC report

From this it would appear that the 374 that were stopped appeared to be connected with offences that are arrestable... Hence my scepticism.

_________________
"Politicians are the same the world over... We build bridges where there aren't any rivers." - Nikita Kruschev

Car crime is such a wide ranging description

Not wishing to get into a pedantic argument with you Pogo, but 32 FPNs were issued and goodness know how many HORT1s, so I would guess the operation had a better ‘success' level than just 14 arrests.

To be honest, I don't have such a problem with this. If the number plate data really is being used for the exclusive purpose of tracking known criminals, insurance evaders etc, and 14 arrests came of it, then it can't be altogether bad.

The worrying bit is how they use that data.

at least someone knows how the system works

_________________
now retired

Yeah... Like a shotgun.

The main thing that alarmed me was that the headline figure was 891 vehicles "of interest to the police". One has to ask why, if so many vehicles were "of interest" yet didn't merit any further action, for what reason are they on the database in the first place?

Oh, of course, silly me... "The milkman said that they were bad 'uns Gov..".

_________________
"Politicians are the same the world over... We build bridges where there aren't any rivers." - Nikita Kruschev

Have to say that having some knowledge of these ANPR opperations through my work for a local authority they do a lot of good. Unlicensed waste carriers (flytippers) tend to get picked up so do untaxed vehicles, and many other disreptable types. No doubt the odd dirty number plate also gets stoped but dealt with by common sense. Generally I do not have a problem with this type of enforcement unlike the cameras.

But aren't one in ten drivers supposed to be uninsured???

Thus a so-called "hit rate" of one and a half percent (covering all manner of offenses) is leaving an awful lot of criminals total freedom to roam?

mb

I’d like to know how many where of interest that they didn’t stop, or couldn’t be bothered stopping because they know they wouldn’t get money out of these type of people. I’ve never seen a traveller’s car taxed, so they would not have insurance, and some probably don’t have driving licences. They are also one of the Majority who flytip.

_________________
Useless laws weaken necessary laws.

Actually, despite starting the thread, I don't have any problems with ANPR per se, my concern was with the accuracy or "inclusion parameters" of the databases used to tag which vehicles were "of interest". They seem to be taking something of a scattergun approach to "intelligence" thus a substantial increase in database size combined with a massive increase in the level of ANPR-based surveillance will inevitably produce an exponential increase in the number of "coincidence-based" data flags. Building a system that automatically "cries 'wolf!'" isn't going to help anyone.

_________________
"Politicians are the same the world over... We build bridges where there aren't any rivers." - Nikita Kruschev

14/391*100 = 3.5% accuracy - why not just stop cars at random - might be cheaper - or sell all the fancy ANPR kit, and use the money to hire some real cops - Durham have some , i believe.

_________________
lets bring sanity back to speed limits.
Drivers are like donkeys -they respond best to a carrot, not a stick .Road safety experts are like Asses - best kept covered up ,or sat on