West Dulwich LTN Action Group - needs your support

[Earl Aelfheah]

Of course… if the data shows an increase in the vehicles counted, it’s an increase. If it shows a decrease in the vehicles counted, it’s because they’re moving slower and also an increase 🙄

Edited September 30 by Earl Aelfheah

[Rockets]

Look Earl, you lifted your rebuttal and your "proof" (almost word for word I hasten to add) on pneumatic counters from the WeArePossible website...so you rail against Twitterati and right-wing groups that you have convinced yourself are at the heart of the opposition to these measures whilst you lift your info from the website of a climate action group....now where is that shrug shoulders emoji....

https://www.wearepossible.org/latest-news/your-ltn-questions

Edited September 30 by Rockets

[Earl Aelfheah]

1 hour ago, Rockets said:

You might think it is nonsense but others do not and the weight of evidence suggests our position is a more reflection of reality than yours.

 

Let's look....

• The manufacturer admitted they are not accurate under slow moving traffic (10kmph) 
• Southwark council stated they were phasing them out and replacing them with more accurate counters
• Even our dear friend Ex-Dulwicher (who works in this sort of stuff) admitted that they are not accurate when used in heavy traffic conditions.

So combine all of the above with the fact that Southwark actively moved monitoring strips closer to choke points (to use their weaknesses to their advantage) then my statement is anything but nonsense.

I don't know where Lambeth have put the monitoring strips in West Dulwich but if they are in areas of slow moving traffic then the pneumatic strips will not be providing an accurate reflection of traffic levels - that is not nonsense, that is a fact.

 

Claims that they can't count vehicles which are travelling less than 10mph aren’t true. You can set a minimum speed default, as well as defaults for the class of vehicle you want to count etc. It all depends on what you're measuring and happens at the software level.

This anti LTN ‘talking point’ started (like many of these things’ with a kernel of truth. In Enfield, there was an error made because they forgot to reset the default speed minimum (which was set to 10mph). This was spotted and the data re-analysed / corrected, but of course, it was used by those looking to undermine any and all data on the impacts of LTNs (a reaction which itself tells you something). 

If you have evidence of the manufacturer saying that they can’t monitor vehicles travelling under 10mph let’s see the sources please.

Edited September 30 by Earl Aelfheah

[raptortruckman69]

Not counting slow cars is typical of the incompetence I'd expect from traffic engineers hired by the council. They have no idea how to use pneumatic counters at all. I'm deeply glad they there are so many upstanding lay citizens who have done their research and posted the results on Twitter and YouTube etc for us to follow. It's good that now anyone is able to correct the so-called professionals.

I shall bring this up next time I meet one of the councilors.

Edited September 30 by raptortruckman69

[Dogkennelhillbilly]

2 hours ago, Rockets said:

the weight of evidence suggests our position is a more reflection of reality than yours.

NB that no evidence or data was cited.

[Rockets]

Here is someone who does know how they work Earl (ExDulwicher - what happened to them I miss them - they were an island of sanity in a sea full of delusion) and look what they say about their accuracy in slow moving and congested traffic....and why Southwark was moving to Vivacity sensors....

 

 

 Rockets said:

Is it a co-incidence that the dashboard numbers are not being updated and many of the tubes seem to have been removed completely after people became more aware of their sub 10km/h limitation? Are the council tryng to mitigate potential exposure?

Ex-Dulwicher said.....I suspect you've actually hit on part of the issue.

Southwark (as with many boroughs in London and cities outside London as well) are moving a lot of their monitoring to Vivacity sensors. They're the camera type things with double lenses you can see on a lot of lampposts around the area and they're vastly more accurate, they can measure pedestrian, bike, car, truck, bus etc very accurately and also measure things like turning flow. They're largely immune to congestion issues and slow moving traffic, or at least can process this as part of the whole package (speed low, flow low, count low = congestion).

 

Edited September 30 by Rockets

[raptortruckman69]

Ha well what's the betting these new sensors will also just mysteriously show a drop in traffic which just so happens to line up with the councils narrative on how marvelous LTNs are? We need to get ahead and start figuring out how it's manipulated before they publish their bogus figures.

[Earl Aelfheah]

Right. So absolutely no evidence evidence of the manufacturer saying that they can’t monitor vehicles travelling under 10mph. Is that perhaps because it's not true?

[Penguin68]

There was no evidence from the  manufactures of cladding used at Grenfell being flammable. Manufactures aren't going to boast of performance issues with their product. Problems with recording of traffic moving under 10mph being recorded accurately in traffic counts using pneumatic detectors is on record. Hence references above to different sensors not having that problem as an argument for their use. 

[Rockets]

https://www.metrocount.com/sites/default/files/2023-05/speedaccuracy.pdf

MetroCount's own installation instructions state:

Site Selection

As with any axle-based classifier, also consider the following:

• Vehicles should be travelling at a constant velocity. Try to avoid bends, intersections and steep inclines.
• Vehicles should be free-flowing. Try to avoid areas with congestion.

 

Now, perhaps Earl you should be asking yourself why the tubes come with a sub 10km/h monitoring filter function, why Metrocount have such a facility and why Enfield had activated it. Why? Because the machines aren't very accurate at monitoring at those speeds. 

 

Now pair that with the fact Southwark actively moved their tubes closer to congestion and it doesn't take a genius to work out why that might be and what games these councils like to play to try and get the results they so desire.

[Earl Aelfheah]

You claimed that:

Quote

....councils were not using those anymore [pneumatic counters] because of the issue of flawed counting or are some still happy to use them because it allows them to manipulate the process?

and that:

Quote

The manufacturer admitted they are not accurate under slow moving traffic (10kmph) 

Metrocounts website states that:

Quote

± 99% Accurate

We're data perfectionists, which is why the RoadPod range includes only sensors that outperform in varied conditions. Independent tests have proven that our car counters lead the industry with ±99% precision. 

You have linked to a document which does not in any way state that they can't count vehicles that travel under 10mph.

This claim has been widely debunked. It is not true.

To state that an increase in vehicle counts is an increase, but that a decrease in vehicle counts is also an increase in traffic is ridiculous. Even you must be able to see just how ridiculous.

Edited October 1 by Earl Aelfheah

[Rockets]

1 hour ago, Earl Aelfheah said:

This claim has been widely debunked. It is not true.

Debunked by who exactly.....?

Perhaps you can explain why Metrocount has a filter for sub-10km/h readings or why they state on their instructions:

• Vehicles should be free-flowing. Try to avoid areas with congestion.

That seems to be counter to your "debunked" argument?

[Earl Aelfheah]

51 minutes ago, Rockets said:

Debunked by who exactly.....?

Perhaps you can explain why Metrocount has a filter for sub-10km/h readings or why they state on their instructions:

• Vehicles should be free-flowing. Try to avoid areas with congestion.

That seems to be counter to your "debunked" argument?

You have pasted a line with no context, and no link. I would love to know how long you had to search to try and find a single sentence which you could misrepresent as supporting your claim.

They don't have a filter for sub-10km/h readings. You can set a lower limit (because you may want to count the number of vehicles travelling over a certain speed). It is defaulted to 10km/h, but can be adjusted. Absolutely no where do they state that they can't measure vehicles travelling at a lower speed.

Their website claims 99% accuracy and links to independent tests from which they draw this conclusion. 

When asked about the false claims that you've repeated, The Times reported that "MetroCount says that their counters are still very accurate for traffic volumes, even in slow and congested traffic conditions, and that accuracy routinely approaches 99% or higher under normal free-flow circumstances".

Your claims that pneumatic counters are wildly inaccurate and count high traffic as low traffic (sounds ridiculous even as I type that), are nonsense.

Edited October 1 by Earl Aelfheah

[Rockets]

We will agree to disagree.....(not for the first time, not for the last time)

 

Search MetroCount installation instructions and you will find the below...it's all publicly available

 

Site Selection

As with any axle-based classifier, also consider the following:

• Vehicles should be travelling at a constant velocity. Try to avoid bends, intersections and steep inclines.
• Vehicles should be free-flowing. Try to avoid areas with congestion.

 

45 minutes ago, Earl Aelfheah said:

Your claims that pneumatic counters are wildly inaccurate and count high traffic as low traffic (sounds ridiculous even as I type that), are nonsense.

Explain to me then why Southwark decided to move the counters near Court Lane to down near Melford Road.....a co-incidence or oversight perhaps....perhaps they weren't reading the instructions supplied by MetroCount!!!

[Earl Aelfheah]

Again, no link to that quote. I suspect it's related specifically to vehicle classifications. Any chance you could share it?

MetroCount says that their counters are "very accurate for traffic volumes, even in slow and congested traffic conditions".

Absolutely nowhere do they say that they do not count vehicles travelling under 10mph. In fact quite the opposite - they claim 99% accuracy.

You can't just state things that aren't true and then say 'agree to disagree'. 

Edited October 2 by Earl Aelfheah

[Rockets]

7 hours ago, Earl Aelfheah said:

Absolutely nowhere do they say that they do not count vehicles travelling under 10mph. In fact quite the opposite - they claim 99% accuracy.

I very much suspect the 99% accuracy applies to "free flowing" traffic - MetroCounts words not mine.

Rest assured when I have time to find the MetroCounter counter manual my paste came from then I will post it. If it does indeed come from a MetroCount installation instruction manual will you stand corrected and issue a grovelling apology? Or will you try to deflect and distract or perhaps construct some reason why they recommend installing in free-flowing traffic and avoiding congestion....?  

[raptortruckman69]

We know traffic has increased. If the counts go up it's because of the increase. If they go down the counts must be wrong because they are obviously not reflecting reality.

 

Everyone knows pneumatic counters can't be trusted and the manufacturers claims of accuracy are wrong.

[Earl Aelfheah]

8 hours ago, Rockets said:

I very much suspect the 99% accuracy applies to "free flowing" traffic - MetroCounts words not mine.

Rest assured when I have time to find the MetroCounter counter manual my paste came from then I will post it. If it does indeed come from a MetroCount installation instruction manual will you stand corrected and issue a grovelling apology? Or will you try to deflect and distract or perhaps construct some reason why they recommend installing in free-flowing traffic and avoiding congestion....?  

The claim that traffic counters don’t work with slow moving vehicles springs from a genuine (and quickly corrected) error in the use of traffic counter software that took place in Enfield. 

The RSU software manual states that:

“The speed filter excludes vehicles outside the specified range. The default range is 10 to 160km/h or 10 to 100 mph. This may need to be changed for sites with an expectation of significant numbers outside of this range”

Enfield failed to change the software’s default setting and so excluded vehicles which were travelling below 10 mph in their analysis of vehicle counts (they would also have missed any travelling over 100). 

When this error was spotted, they were able to easily re-analyse the data by changing the lower limit and running the reports again through the software. 

Note: this increased the number of vehicles counted.

Inevitably this was grabbed on to by those who had no interest in the reanalysed counts, instead using it to make the following claims:

• that the oversight was deliberate and had been used as a tactic to show drops where there were none and to hide or minimise increases in traffic
• that ATCs cannot count slow vehicles
• that where the same technology showed increases at some sites using the same methods, that this was accurate and should be given significant weight

Obviously the three assertions are totally incompatible with each other / logically inconsistent. But when you’re just looking for ways to prove your point, it doesn’t really matter – it’s just about kicking up dust (a tactic regularly and cynically on display across this thread).

To answer your question, no, I don't care whether you manage to track down a recommendation on the ideal conditions for installing ATCs. Nor am I demanding (like you) a 'grovelling apology". I'm merely asking for a correction. You claimed that:

“The manufacturer admitted they are not accurate under slow moving traffic (10kmph) ”

Yet MetroCount when asked about the Enfield 'controversy' by the Times, stated on the record that their counters are:

“very accurate for traffic volumes even in very slow and congested traffic conditions”

Edited October 2 by Earl Aelfheah

[Dogkennelhillbilly]

7 hours ago, raptortruckman69 said:

We know traffic has increased.

NB no data or evidence cited.

[Rockets]

Earl, you are selectively clipping elements of the Time article.

 

here is the full text for reference where MetroCount actually admit the fallibility to the Times. https://www.thetimes.com/uk/politics/article/flaw-in-roadside-counters-for-low-traffic-schemes-j6wbwvzjn

 

 

 

Traffic counters used to monitor the impact of low traffic neighbourhoods (LTNs) are not accurately recording vehicles during heavy congestion, The Times can reveal.

Research suggests that between 5 and 35 per cent of cars, vans and lorries in slow moving or stop-start traffic are not being counted, calling into question the claimed success of the controversial schemes.

Scores of LTNs were introduced during the pandemic and many more are planned. They use bollards, planters or camera enforcement to block through traffic in residential areas. The idea is to encourage people to walk or cycle instead of using their cars for short journeys. However, critics say the schemes force traffic on to a small number of surrounding roads, increasing congestion and pollution.

Councils that have introduced LTNs have mostly hailed them as a success, pointing to data showing they have cut traffic both inside the areas and on some of their boundary roads. However, local residents have often been baffled by these claims, saying they have witnessed significant increases in congestion on the boundary roads.

The company says the counters are “not designed to work” in stop-start traffic and are recommended to be used in “free-flowing conditions”.

It explained: “Vehicles travelling very slowly might not be classified correctly, either the axle hits are too far apart so it splits them and places them into an unknown vehicle class, which doesn’t get included by default, or it attaches those axle hits to a vehicle in front or behind.”

This means if there is little or no congestion at the measuring points before the LTN, the number of vehicles counted is likely to be accurate. However, if the LTN creates congestion at the count points, the post-implementation surveys will not report the true number of vehicles.

There are also fears some counters may not have recorded vehicles travelling under 6.2 mph at all. One north London council has already been forced to admit it under-reported congestion on the boundary roads surrounding one LTN.

Enfield council confessed that cars, vans and lorries travelling in congestion had not been recorded after the Fox Lane LTN was introduced because a software update had changed the setting without its traffic engineers realising. MetroCount said the council’s contractor appeared to have made “a deliberate choice to change the default setting, contrary to MetroCount documentation that advises caution when surveying slow-moving or congested traffic”.

 

It added: “Every single report produced by our software lists all the parameters selected at the top of the report, including any speed range settings.”

The Times approached eight inner London councils that introduced LTNs during the pandemic to ask what settings they had used on their counters, whether they had been adjusted after their schemes were implemented and whether they were confident in the accuracy of their data, but none of them answered. Almost all of the councils have presented their LTNs as a success, claiming traffic reductions even on some boundary roads.

Enfield council conducted a manual count of vehicles on four roads that also had automatic counters. This survey found that the automatic counters under-recorded nearly 3,000 vehicles — the equivalent of 5.4 per cent — over a 12-hour period.

Dozens of videos have emerged on social media of cars moving very slowly or being stationary over counters on LTNs’ heavily congested boundary roads.

One resident of Enfield was so suspicious of the data reported by the council on his heavily congested road near the edge of an LTN that he trawled through 24 hours of CCTV to see how many vehicles had actually passed his home.

 

Ediz Mevlit, a bus driver from Palmers Green, said: “The council said only 1,845 cars a day passed through my road on average and congestion had reduced. But when I watched the CCTV back, I counted 2,523, that’s about 30 per cent more — and I probably missed a few because I sped up the footage. I was so angry because they had been making me feel paranoid. It’s the gaslighting, telling me traffic has reduced when it hasn’t.”

After Enfield council’s recording error was discovered, it re-ran all the data without the 6.2 mph filter and recovered some of the undercounted vehicles. This changed its reported data from a 5.7 per cent increase on boundary roads to 8 per cent but critics say new data still did not record vehicles during heavy congestion because of the counters’ inherent limitations.

In a report on Hackney council’s LTNs, John Wilde, a director at Charles & Associates Consulting Engineers, said: “Automatic traffic count surveys cannot be considered as broadly accurate [on congested roads]. A CCTV method survey would be more robust, and would also capture the stationary or slow-moving traffic conditions, whilst also allowing for clearer assessment of the peak periods.”

Automatic counters have also been the primary source of data used by academic studies suggesting that LTNs work. This week a petition demanding the government carry out an independent review into LTNs surpassed 10,000 signatures, forcing a response. It said the Department for Transport had already appointed the University of Westminster to “undertake an independent evaluation of active travel schemes funded in 2020-21”.

The director of the University of Westminster’s Active Travel Academy is Professor Rachel Aldred, a former trustee of the London Cycling Campaign, which has been one of the most vociferous advocates of LTNs.

Enfield council said new traffic data on its boundary roads meant there was “no material change” to its previous conclusion that the Fox Lane LTN should be kept. It added that it did not rely on automatic counters alone and used other methods of assessment such as bus journey times to consider the impact of its schemes.

The council also said it was not appropriate to compare directly the data between the manual counts and the automatic count, or draw conclusions from the disparity because they were not conducted at exactly the same spots on the roads.

MetroCount says that its tube-based counters are still very accurate for traffic volumes, even under very slow and congested traffic conditions because the axles passing are continuously recording. It says that in most circumstances this would exceed 95 per cent accuracy but that the reporting of volumes under slow-moving conditions requires users to change the default settings.

In a statement it added: “Our standard settings are recommended in normal free-flow circumstances and when also wanting to display very accurate speeds, classes, headway, gaps and other higher level information in addition to the basic traffic volume. Under these conditions accuracy routinely approaches 99 per cent or higher, as verified by many departments of transport globally.

“All automatic counting technologies will have some limitations in very high congestion, and even manual and video counting is known to have accuracy issues for long-term counting. It is incumbent upon users to know how, where and when to best install traffic counters and use complex associated software.”

The residents feeling the impact of LTNs

It was a struggle for Christiane Comins to get to the protest outside Islington town hall this week but she got out her crutches and gritted her teeth. Comins, 53, has multiple sclerosis (Jack Malvern writes).

While she has a blue badge that gives her permission to drive through her neighbourhood, this is of no use for the delivery drivers she needs for essentials or the friends she relies on for help.

Her neighbourhood, in the Barnsbury area of Islington, is the latest in the north London borough to be proposed as a low-traffic neighbourhood (LTN). It prompted a protest of more than 100 residents outside the town hall on Tuesday.

“Because I’m disabled, I’m not only reliant on taxi journeys to or from the hospital but also reliant on friends coming to visit me,” she said. “I’m reliant on food deliveries to the house. They won’t be able to come. I may just drown in a pool of my own tears. I used to say I should jump in the Thames, but I can’t climb the bridge any more.”

She said her MS comes and goes and that she dreads returning to a bedridden state. The last time, her friends were able to visit and even hold a party for her. “That can’t happen now if Barnsbury LTN goes ahead. I have MS and it’s a bugger. I go up and down.

“Of course I support cyclists, but do they need every street in the neighbourhood? No, they don’t.” She worries not only for other people with disabilities but for local shops.

“I worry a lot for some businesses like our butcher, who is feeling the pinch from [existing] LTNs. It’s quite frightening how little say you have.” Others at the protest complained that their daughters felt unsafe walking home at night through streets deserted of traffic.

Nicholas Mason, 78, a retired solicitor, said his journey time to pick up his granddaughter from school in Tottenham, north London, had tripled. Jonathan Harrison, 75, a retired architect, added that the zones were frustrating his journeys to hospital for cancer treatment.

A few weeks ago he received £480 in traffic fines for driving past a camera near his home. “I never saw a sign saying don’t come in this road. I had no idea they’d closed it. I think it’s totally outrageous. There’s been no consultation. None of this was in their manifesto, that they’re going to make it impossible to drive. It’s a pointless, ridiculous, vengeful policy.”

[Earl Aelfheah]

Wow, that's a very long winded self own. Well done.

So in summary, you've stated that

“The manufacturer admitted they are not accurate under slow moving traffic (10kmph) ”

..and then in defence of this claim, quoted the manufacturer:

"MetroCount says that its tube-based counters are still very accurate for traffic volumes, even under very slow and congested traffic conditions because the axles passing are continuously recording. It says that in most circumstances this would exceed 95 per cent accuracy"

 

Edited October 2 by Earl Aelfheah

[Rockets]

42 minutes ago, Rockets said:

The company says the counters are “not designed to work” in stop-start traffic and are recommended to be used in “free-flowing conditions”.

It explained: “Vehicles travelling very slowly might not be classified correctly, either the axle hits are too far apart so it splits them and places them into an unknown vehicle class, which doesn’t get included by default, or it attaches those axle hits to a vehicle in front or behind.”

Just leaving this here Earl.....

Even Rachel Aldred questions the accuracy of tubes in a report she did - I doubt she does this if the 99% accuracy you claim is correct. I very much suspect 99% applies to those placed in free-flowing traffic. I very much suspect you have to set them to either monitor traffic above 10km/h or set it to monitor under 10km/h - which is why they are supplied with default setting over 10km/h.

 

Rachel Aldred said in this report:

https://westminsterresearch.westminster.ac.uk/download/74b26baccb2dbc0d26f1ca1773b3cdcd08402ef0e79fae1908f79d77c2cb7653/6168872/1-s2.0-S2213624X23001785-main.pdf

There are known issues with data quality. Usually, reports used Automatic Traffic Counters (ATCs) to monitor traffic, in most cases ‘tubes’ across the road. These are imperfect. Parked or very slow-moving motor traffic may affect results; although in most cases, count sites were placed away from junctions where queueing is likely, which should reduce this problem.

 

And remember...Southwark moved monitoring strips CLOSER to junctions and congestion.......perhaps they did not read the instructions properly.....

[Earl Aelfheah]

2 minutes ago, Rockets said:

Just leaving this here Earl.....

So exactly as I suspected originally, you (and rather misleadingly the Times article) have conflated classifications of vehicle type with volume / count data.

MetroCount have explicitly stated that when it comes to volumes that ATCs are 

"very accurate for traffic volumes, even under very slow and congested traffic conditions"

They state that for traffic volumes, accuracy exceeds 95%. This does not remotely fit with your statement that “The manufacturer admitted they are not accurate under slow moving traffic (10kmph) ”

[Rockets]

Earl, for goodness sake read the whole article...MetroCount say the below......I can't spell it out any more simply...

 

If even Rachel Aldred is calling the accuracy under slow moving vehicles into question you seem to be on an island of 1 with your position.

 

The company says the counters are “not designed to work” in stop-start traffic and are recommended to be used in “free-flowing conditions”.

It explained: “Vehicles travelling very slowly might not be classified correctly, either the axle hits are too far apart so it splits them and places them into an unknown vehicle class, which doesn’t get included by default, or it attaches those axle hits to a vehicle in front or behind.”

[Earl Aelfheah]

19 hours ago, Earl Aelfheah said:

Again, no link to that quote. I suspect it's related specifically to vehicle classifications. Any chance you could share it?

MetroCount says that their counters are "very accurate for traffic volumes, even in slow and congested traffic conditions".

Absolutely nowhere do they say that they do not count vehicles travelling under 10mph.

So exactly as I suspected. You are talking about vehicle classifications (that is determining the type of motor vehicle):

“Vehicles travelling very slowly might not be classified correctly"

MetroCount have explicitly stated that when it comes to volumes (i.e. the count of the number of motor vehicles) that ATCs are 

"very accurate for traffic volumes, even under very slow and congested traffic conditions"

They state that for traffic volumes, accuracy exceeds 95%. This is completely at odds with what you have claimed the manufacturer has said. 

Edited October 2 by Earl Aelfheah