All hail the new AI speed camera

[texaxile]

Well, if its good enough for Jimmy Savile then it's good enough for.....oh wait.....

https://www.youtube.com/watch?v=qnZ6tOVQwLk

No idea why people don't want to wear seatbelts. I was very surprised to see the amount of fine for non compliance though. I also didn't realise it was endorseable now. Every day's a schoolday on PH.

[Simpo Two]

vonhosen said:

Simpo Two said:

vonhosen said:

FMOB said:

I don't condone driving without a seat belt on or using a mobile phone but adding AI means the seat belt and phone are probably just the start of what could be detected using a camera. A clean licence is rapidly becoming an endangered species.

Rubbish.

DVLA figures show that about 95% of licence holders have zero penalty points.

Points per mile travelled would be more representative. Many people will have a licence but drive rarely if at all.

Anyway, I shall be immune from two offences as I always wear a seat belt and hate mobile phones so even if it's with me, it's switched off in the boot.

But what about bananas eh? Is it an offence to eat a banana, and how good is the AI camera at identifying fruit varieties? If it has a mass spectrometer it could fine you for farting too.

He made the claim about licences.
I'm merely point out the facts which show that his hunch was incorrect & pointed him towards the data.

Data is a wonderful thing. But my hunch is that if you go out tomorrow and flag down 20 cars, more than one driver will have points on his licence. That would tell us if theory is reality.

[FMOB]

Simpo Two said:

vonhosen said:

Simpo Two said:

vonhosen said:

FMOB said:

I don't condone driving without a seat belt on or using a mobile phone but adding AI means the seat belt and phone are probably just the start of what could be detected using a camera. A clean licence is rapidly becoming an endangered species.

Rubbish.

DVLA figures show that about 95% of licence holders have zero penalty points.

Points per mile travelled would be more representative. Many people will have a licence but drive rarely if at all.

Anyway, I shall be immune from two offences as I always wear a seat belt and hate mobile phones so even if it's with me, it's switched off in the boot.

But what about bananas eh? Is it an offence to eat a banana, and how good is the AI camera at identifying fruit varieties? If it has a mass spectrometer it could fine you for farting too.

He made the claim about licences.
I'm merely point out the facts which show that his hunch was incorrect & pointed him towards the data.

Data is a wonderful thing. But my hunch is that if you go out tomorrow and flag down 20 cars, more than one driver will have points on his licence. That would tell us if theory is reality.

Maybe points dished out per annum is a better metric along with speed awareness course per annum. It is easier than ever to pick up penalty points when out driving these days and is only going to become more likely as AI gets connected to more cameras and sensors.

Maybe the discussion should be about raising the points limit for a ban under the totting up procedure from 12 to something higher but then again the 12 point limit doesn't apply equally to all as we sometimes see in the media.

[vonhosen]

Simpo Two said:

vonhosen said:

Simpo Two said:

vonhosen said:

FMOB said:

I don't condone driving without a seat belt on or using a mobile phone but adding AI means the seat belt and phone are probably just the start of what could be detected using a camera. A clean licence is rapidly becoming an endangered species.

Rubbish.

DVLA figures show that about 95% of licence holders have zero penalty points.

Points per mile travelled would be more representative. Many people will have a licence but drive rarely if at all.

Anyway, I shall be immune from two offences as I always wear a seat belt and hate mobile phones so even if it's with me, it's switched off in the boot.

But what about bananas eh? Is it an offence to eat a banana, and how good is the AI camera at identifying fruit varieties? If it has a mass spectrometer it could fine you for farting too.

He made the claim about licences.
I'm merely point out the facts which show that his hunch was incorrect & pointed him towards the data.

Data is a wonderful thing. But my hunch is that if you go out tomorrow and flag down 20 cars, more than one driver will have points on his licence. That would tell us if theory is reality.

As the data shows that would depend on things such as where you do it.
It doesn't alter the facts that about 95% of licence holders have zero points (which isn't a theory).

[vonhosen]

FMOB said:

Maybe points dished out per annum is a better metric along with speed awareness course per annum. It is easier than ever to pick up penalty points when out driving these days and is only going to become more likely as AI gets connected to more cameras and sensors.

It's even easier, by some magnitude, to offend & go completely undetected.
How many times have you offended in your life?
How many times have you been caught & picked up penalty points?

[Nibbles_bits]

Geffg said:

Yes using a phone, no seatbelt, speeding etc is illegal and one death through either is one too many, the annoying thing is for the last god knows how many years all you hear about is camera for this camera for that, gonna clamp down on motorist for this, parking clampdown, charge more for parking, road tax going up because we’re the devil and the main cause of global warming, can’t drive down this road or that road, bus lanes, put obstructions in roads that are dangerous, charge for parking outside your own house, etc etc.
Yet crime is on the up, drug related crime / deaths, stabbing, shootings etc are on the rise. Wouldn’t it be better use of resource and money to concentrate on sorting those issues out as I’m sure there must be more deaths and injuries from that than motorists cause. But as usual we are mostly legit and so a very easy target to make money from.
Watching some of the cop programmes it would be better to drive round on false plates as the penalty for that seems to be less than being caught doing something whilst all legit.

The fines are issued by the Police on behalf of the Council.

The amount of police resources that go into this is minimal and would have no impact on crime if they didn't issue the tickets.

[I am alright Jack]

texaxile said:

Well, if its good enough for Jimmy Savile then it's good enough for.....oh wait.....

https://www.youtube.com/watch?v=qnZ6tOVQwLk

No idea why people don't want to wear seatbelts. I was very surprised to see the amount of fine for non compliance though. I also didn't realise it was endorseable now. Every day's a schoolday on PH.

agtlaw said:

Seatbelt offence is non-endorsable. Fixed penalty £100. If prosecuted then it’s a level 2 offence (up to £500).

[Dave Finney]

vonhosen said:

It's even easier, by some magnitude, to offend & go completely undetected.
How many times have you offended in your life?
How many times have you been caught & picked up penalty points?

But what REALLY matters is:

How many times have you crashed.
How many of your crashes resulted from you committing an offence?

We do know the answers.
https://speedcamerareport.co.uk/speeding/

[Countdown]

Dave Finney said:

But what REALLY matters is:

How many times have you crashed.
How many of your crashes resulted from you committing an offence?

We do know the answers.
https://speedcamerareport.co.uk/speeding/

is that because speed cameras forces most people to comply with the limits (or at least not go totally bananas).

It's a bit like saying hardly anybody dies in lift accidents and therefore we don't need such stringent LOLER regulations. i.e it might well be that only 14% of fatal collisions involve speeding because most people don't speed because of cameras.

Actually it would be good if we could compare KSI stats between the Uk and somewhere where cameras aren't used.

[Dave Finney]

Countdown said:

is that because speed cameras forces most people to comply with the limits (or at least not go totally bananas).

It's a bit like saying hardly anybody dies in lift accidents and therefore we don't need such stringent LOLER regulations. i.e it might well be that only 14% of fatal collisions involve speeding because most people don't speed because of cameras.

Actually it would be good if we could compare KSI stats between the Uk and somewhere where cameras aren't used.

The short answer is no.
What you're asking is:
"Are there any area-wide benefits due to speed cameras?"

In the early days, they really did think there would be.
The DfTs first national report stated:

"All areas joining the system should endeavour to measure the impact the system [deploying speed cameras] is actually having on the NHS,
for example by monitoring the number of hospital bed-days required for people injured in road traffic accidents."

In their final national report, the DfT realised that there were no area-wide benefits at all (road safety improvements had actually got worse).
https://speedcamerareport.co.uk/official-reports/

[Countdown]

Dave Finney said:

The short answer is no.
What you're asking is:
"Are there any area-wide benefits due to speed cameras?"

In the early days, they really did think there would be.
The DfTs first national report stated:

"All areas joining the system should endeavour to measure the impact the system [deploying speed cameras] is actually having on the NHS,
for example by monitoring the number of hospital bed-days required for people injured in road traffic accidents."

In their final national report, the DfT realised that there were no area-wide benefits at all (road safety improvements had actually got worse).
https://speedcamerareport.co.uk/official-reports/

Nothing that complicated Just a comparison of KSI in countries or areas which HAVE speed cameras against countries/areas which don't have speed cameras.

[Dave Finney]

Countdown said:

Nothing that complicated Just a comparison of KSI in countries or areas which HAVE speed cameras against countries/areas which don't have speed cameras.

That is what they did.
Areas with cameras compared with areas without.
They found speed cameras provided no area-wide benefits.

[Countdown]

Dave Finney said:

That is what they did.
Areas with cameras compared with areas without.
They found speed cameras provided no area-wide benefits.

Apologies, I should have read the link that you posted

SpeedCameraReport said:

The 4YE is the largest and most comprehensive speed camera report in Britain but it is a very deceptive report.
The conclusions in the summary are not supported by the evidence within the report. The Executive summary (p2):

“Overall 42% fewer people were killed or seriously injured. At camera sites, there was also a reduction of over 100 fatalities per annum (32% fewer). There were 1,745 fewer people killed or seriously injured…”
But the authors knew that these reductions were not caused by the speed cameras.
Near the end of the report (from p141) there is an estimate of the effect of site selection (they call it RTM, or regression to the mean). It says (p155):

“Thus RTM accounts for about three fifths of the observed reduction in fatal or serious collisions with the effects of the cameras and trend each accounting for a fifth“.
So most of the reductions would have occurred anyway without the cameras, and the cameras only caused a relatively small further reduction. But …

My interpretation of that was

- Huge reductions in KSI
- But 60% of that is RTM

Doesn't that suggest that the cameras contributed towards the remaining 40%?

To be honest i was thinking more widescale. I stick to the speed limits all the time rather than constantly worry about speed cameras and I'm sure a lot of other people do too. So wouldn't the average fall in speed contribute towards a reduction in KSI (not just at the camera sites but on a wider scale?)

[vonhosen]

Dave Finney said:

vonhosen said:

It's even easier, by some magnitude, to offend & go completely undetected.
How many times have you offended in your life?
How many times have you been caught & picked up penalty points?

But what REALLY matters is:

How many times have you crashed.
How many of your crashes resulted from you committing an offence?

We do know the answers.
https://speedcamerareport.co.uk/speeding/

You don't only get prosecuted when you crash.
You can get prosecuted when you offend & get caught.
If a clean licence matters to you, not getting caught offending matters.
As I say though, despite what others may say here, there are comparatively few prosecutions when compared to the amount of offending.

Limits do not exist solely to stop crashes, they exist for many reasons.

Edited by vonhosen on Monday 26th February 17:24

[FiF]

NikBartlett said:

I think some fun can be had fooling the AI into thinking you have commited an offence

Some years back found it more comfortable when wearing a gilet to keep that unbuttoned and then it draped over the correctly fastened seatbelt. Something to do with the route of the belt pulling on the hem edge.

Got stopped for no belt but fortunately the officers realised their mistake straightaway and we had a laugh and a reasonable discussion.

Don't think such a discussion would be possible with a stupid camera or back office, never mind Artificial (Un)Intelligence.

[gavgavgav]

Sebring440 said:

LivLL said:

Interesting to read that 1/3rd of people killed in road collisions in 2021 were not wearing a seat belt!

Aha! So two-thirds of people killed in road collisions in 2021, were wearing a seat belt?

Get those seat belts off!

How many were riding a motorbike?

[Dave Finney]

Countdown said:

My interpretation of that was
- Huge reductions in KSI
- But 60% of that is RTM
Doesn't that suggest that the cameras contributed towards the remaining 40%?

To be honest i was thinking more widescale. I stick to the speed limits all the time rather than constantly worry about speed cameras and I'm sure a lot of other people do too. So wouldn't the average fall in speed contribute towards a reduction in KSI (not just at the camera sites but on a wider scale?)

Yes, it would seem reasonable that the average fall in speed might contribute towards a reduction in KSI on a wider scale,
but the evidence shows that it does not.

In the early days, the DfT thought that cameras would have:
1. effects at the camera sites,
2. area wide effects, not just at the sites.

We can see this in their 1st report.
Huge KSI reductions at the sites, so they assumed the cameras caused those reductions,
and they stated the need to evaluate area wide effects for their next report.

In their final report they had compared areas with and without cameras and found no benefit,
and checked hospital beds and found no benefit.
So this was excluded from their report.

At the camera sites, they claim that their RTM estimate gave a result showing that the cameras caused 20% of the KSI reduction at the sites.
But they didn't mention that in the summary at the top,
and did not adjust their figures to reflect that.

Worse still,
When we look at the method they used to evaluate RTM,
we find that their model estimate found that the cameras had caused an increase in KSI at their sites.
So what they did was just add collisions to their model estimate until it produced the opposite result!

The standard of evidence in official speed camera reports is shockingly bad,
so bad that it would be illegal in any other area of public safety!

[Sporky]

Does that mean speed cameras reduce collisions? Or did I miss something?

[Dave Finney]

Sporky said:

Does that mean speed cameras reduce collisions?

The short answer is no.
The evidence suggests that speed cameras increase fatal and serious collisions at their sites.

A longer answer is:
There is an effect caused by how camera sites are selected.
It's called RTM.
No official report has ever removed RTM.
They thought that to do that was impossible.
The effect of speed cameras has therefore never been officially established.

I developed the FTP method.
It is the only method that can remove RTM, given the data available.
https://speedcamerareport.co.uk/rtm-regression-to-...
I evaluated mobile speed cameras in TV and found an increase fatal and serious collisions at their sites.
My report had completely removed the effect of RTM.

It was a world first.

Since then, 3 more reports have used my FTP method accurately:
Fixed cameras in TV,
Fixed cameras all across London,
All fixed and mobile cameras in Wales.

All found increases in fatal and serious collisions at the speed camera sites.

[Sporky]

Sorry, being thick - what are KSI, RTM, and FTP in this context?