Dear Mr Belchamber
Thank you for your
two further emails on the above subject.
I am unable to add
much more to my previous responses to you.
I would reiterate
that the need for the signals at Fleetsbridge was established in 1993. They were
installed by Dorset County Council as Highway Authority at that time, as part of
a major linked scheme including the Tesco access and removal of the small
roundabout at the Waterloo Road/Cabot Lane junction. The need was established
using traffic modelling and transport planning techniques at the time by Dorset.
When the Borough became the Highway Authority in 1997 the junction was already
well established as a signalised roundabout, albeit working part-time.
I believe that I
have already given comprehensive, reasonable and considered responses to your
previous emails. It is evident from your latest reply that you neither accept
these responses or the explanation contained therein and you will, I suspect,
come back with further very similar queries which will again not add anything
further to the debate.
I cannot really see
any benefit in carrying on this discussion further. I have spent significant
public money on dealing with your continued requests on this subject and in this
respect there is indeed a limit. I do not have the time to keep responding to
such requests just to satisfy your curiosity, or to serve your personal
campaigning, when I have repeatedly said that the decision on future operation
will be made by Members and not you.
It is clear from
your last comment of the 9 December email that you threaten to circulate
previous (or any further responses) in order to discredit both the organisation
and me. I therefore have no further comments to make on this and I will not
respond to any further emails from you on this particular topic.
Kind regards
Martin Baker
Senior Engineer (Road Safety and Urban Traffic Control)
T:
01202 262073
F:
01202 262091
E:
m.baker@poole.gov.uk
W:
www.boroughofpoole.com
Transportation Services, Borough of Poole, St Johns House, Serpentine Road,
Poole BH15 2DX
www.dorsetroadsafe.org.uk
From:
Ian Belchamber (gmail) [mailto:ianbelchamber@gmail.com]
Sent: 12 December 2011 08:10
To: Martin Baker
Subject: Re: Fleetsbridge Roundabout Traffic lights - a proper analysis -
traffic lights probably INCREASED accidents -POOLE COUNCIL PLEASE RESPOND
There's one
more important point to add:
9. Dorset Road
Safe (and previously DSCP) have made bold claims about reducing accidents and
casualties in Dorset over the years. By claiming that any reduction at
Fleetsbridge is ONLY due to the traffic lights, and not due to anything else
including DRS, are you effectively agreeing with me that their claims are bogus?
Sent: Friday,
December 09, 2011 2:55 PM
Subject: Re:
Fleetsbridge Roundabout Traffic lights - a proper analysis - traffic lights
probably INCREASED accidents -POOLE COUNCIL PLEASE RESPOND
Many thanks
again for another proper response, I too (although it might not seem like it)
have limited time.
So I'll be as
brief as possible with some new points and I'd ask for at least one more
response on this:
1. There were 2
"roundings", one of yours and a smaller one "for publication" resulting in
inflation of a number from £266,448 to £300,000. Rounding is appropriate for
clarity but this is pushing it. I also pointed out that the remaining figure was
being vastly inflated (about another £100,000) by the erroneous "cost of
death". If there was a proper analysis / justification done in advance, why was
this not already available (perhaps on the council web site) to refer people to
in the Echo item?
2. Your whole
statistical analysis seems to be comparing part time with full time. As I said
it does seem that part time lights increased accidents and maybe there's
something that can be learned from this but it is irrelevant to this discussion.
We want to see why we have lights at all, compared to no lights. (Obviously,
part time lights would cost the same as full time lights).
3. I don't see
what "response time" has got to do with it. The council should have had a
properly thought out proposal / justification (available for the public to
respond to) before the work was done or even planned.
4. If proper
reasoning "backs anyone into a corner" then those that it disadvantages have not
done enough of it.
5. "The need for the signals had already been well established". No it has not.
That is what this is about.
6. "The figures quoted were attributable to the success the
signals had demonstrably achieved in reducing and maintaining lower accident
levels" I say again, please can you detail what you believe the accident
reduction to be, no lights to full time lights, and how you have come to
this value. This cannot ignore the factors I'll list below. This has NOT been
demonstrated yet, other than the simplistic comparison of 7 (actually 6.6)
accidents a year 1987 to 1993 and and average of 2.7 from 2000 to 2011, which is
not scientifically credible for at least the following reasons which I repeat:
-
in the direct period after full time lights were installed (a better period for
comparison but shorter) accidents went UP from 6.6 to 8.8
-
2000 to 2011 shows downwards trend which could not have been due to the lights
but helps to make the "full time" average number smaller.
-
changes in reporting, road markings, long term trends, etc. have not been
considered
- there is 7
years between these periods when all kinds of other things have shifted
- if you "idiot
proof" one section of road many of the accidents that would have happened there
will probably just happen somewhere else, they are not reduced
- the counts
1987 to 1993 are very variable and do not allow the norm to be estimated with
high confidence
- no value has
been placed on increasing congestion, pollution, cost, journey times, stress and
frustration (and therefore danger) to thousands of drivers every day who are
unnecessarily delayed by lights, here (and elsewhere, Mountbatten, etc).
7. The lights
should never have been installed in 1993. When they came to the end of their
life, this should have been obvious and they should have been simply removed,
and the money better spent on something else (perhaps a proper traffic cop for a
year). I'm quite certain a proper traffic cop could save many lives a year, not
just less than a handful of minor scrapes, and just for clarity, a jobsworth
sitting in a van with a speed camera (or in an office sending out fines) is NOT
a proper traffic cop.
8. If what I
have expressed is clear and stands up to reasoning, then if it conflicts with
what has been done, it should not be ignored by you, the appropriate committee
or anyone else. I have no powers either but when I see something wrong I will
try to do something about it.
Therefore, I
expect the points I have made to be properly answered. If my interest in this
results in greater care being taken and better decisions in the future, I will
be very pleased. Most would expect that there has in fact been a diligent and
proper process, in which case answers to my concerns will already be available
in which case please let me know. If I am wrong on any then I will be happy to
accept it, but there must be CREDIBLE and complete answers. If you can't answer
these points I will seek answers from the usual list (and maybe others).
Sent: Friday,
December 09, 2011 12:25 PM
Subject: RE:
Fleetsbridge Roundabout Traffic lights - a proper analysis - traffic lights
probably INCREASED accidents -POOLE COUNCIL PLEASE RESPOND
Dear Mr Belchamber
Thank you for your
email and also to agreeing to my request to only direct your replies to me.
As I have mentioned
in previous emails my resources are very limited and so it is not possible to
respond to all the points you have raised, however I would (as briefly as time
permits) reply as follows:
I note what you
have said about the Echo article, but we have no control over what the Echo
finally print when it goes to press. Certainly this is one occasion where I was
asked to supply background information but only a much-edited version was
eventually printed. The rounding up or down of certain figures is not something
I am keen on but it makes it easier for the casual lay reader to digest –
certainly there were a lot of subsequent posts on the website asking how there
can possibly be 2.6 accidents a year but as you are fully aware this is a
consequence of calculating averages.
With regards to
demonstrating statistical significance I would suggest that you use the chart
supplied by TMS consultancy. TMS do all the road safety engineering training on
behalf of RoSPA. Please download the following spreadsheet:
http://www.tmsconsultancy.co.uk/resources/analysis-tools/statistics-tests-fyrr
Once open look up
the ‘All Purpose Poisson Test’ tab along the bottom. If you then enter accident
numbers for whole years into the boxes provided – you can enter up to ten years
maximum – it runs a significance/confidence test. You are of course free to
enter any of the data I have previously supplied you in any combination, but if
you add in the 5 years before and after full time signals using the following
numbers as an example you will see how I remain confident about their
effectiveness in accident reduction terms.
|
Fleetsbridge
|
|
|
|
|
|
|
|
|
|
|
1996
|
8
|
|
|
|
|
1997
|
17
|
|
|
|
|
1998
|
14
|
|
Part Time Signals
|
|
1999
|
9
|
|
|
|
|
2000
|
12
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2001
|
2
|
|
|
|
|
2002
|
5
|
|
|
|
|
2003
|
2
|
|
Full Time Signals
|
|
2004
|
5
|
|
|
|
|
2005
|
3
|
|
|
|
I disagree that we
had to ‘create a case’ in response to an Echo query but would repeat that
because response times are very tight due to publication deadlines things can
get lost in translation at times.
I can see why you
may be continuing with your request for more detail but I do feel that you may
be trying to back us into a particular corner with your reasoning. The decision
to turn on the signals back on full time was agreed back in 1999/2000. The need
for the signals had already been well established. The recent refurbishment of
the junction is purely related to the need to replace poles that were beginning
to structurally fail and were becoming unsafe. The refurbishment scheme is a
revenue replacement and is not a road safety scheme in its own right. The
figures quoted were attributable to the success the signals had demonstrably
achieved in reducing and maintaining lower accident levels and were not intended
to justify the refurbishment scheme.
Whenever we
refurbish signals (be it junctions, puffins, toucans etc.) we do so ultimately
because they have reached the end of their design life. The need for the
facility is usually well established, whether it be a safe crossing point for
pedestrians, as part of a safety scheme and/or to balance traffic flows and
assist vulnerable road users. We have removed signals in the past but this has
primarily been due to changes in road layouts etc.
I can see that you
have fully and very clearly expressed your opinion on the need or otherwise for
the signals, however as I said in my previous reply this need will ultimately be
decided by Members and not in any email forum between the two of us.
I certainly have no
powers to remove or change things here without prior authority from the
appropriate Committee.
Kind regards
Martin Baker
Senior Engineer (Road Safety and Urban Traffic Control)
T:
01202 262073
F:
01202 262091
E:
m.baker@poole.gov.uk
W:
www.boroughofpoole.com
Transportation Services, Borough of Poole, St Johns House, Serpentine Road,
Poole BH15 2DX
www.dorsetroadsafe.org.uk
From:
Ian Belchamber (gmail) [mailto:ianbelchamber@gmail.com]
Sent: 07 December 2011 20:56
To: Martin Baker
Subject: Re: Fleetsbridge Roundabout Traffic lights - a proper analysis -
traffic lights probably INCREASED accidents -POOLE COUNCIL PLEASE RESPOND
(If you
received any earlier responses please discard them, I may have sent in error
before complete).
Thank you so
much for preparing a proper response. One of the greatest problems about this
(and indeed many things the council does) is the poor level of communication. I
think you can understand that the response given by Steve Tite at the time (an
average of 7 before and 2.6 in a period 7 years later, therefore a £300,000
saving per year) was totally inadequate and did indeed indicate a lack of care
and knowledge. Even now I get the impression that a proper case has had to be
created after the work, rather than before the work when it should have been
done anyway to properly establish need, and would have been available to put to
anyone concerned as soon as required.
I'd like to
answer your points:
1. Based on the
analysis you have done, can you detail what you now believe the reduction due to
the lights to be, and how you arrive at this result? In the 5 years before, the
variation (8) exceeds the average (6.6), there is too much variability from
these 5 points to deduce anything. We would have to see data going back perhaps
10 years to know where the average is. The 10 years after 2001 do appear to have
a lower average, but with a downward trend - clearly the lights could not be
responsible for the downward trend. So it would be more realistic to use an
average as it would have been for around 2001 for the "after" value, which would
be about 4 per year. "It is possible to deduce from the figures previously
supplied to you that the reduction achieved is statistically significant" - yes
I'm sure, but the point is, achieved by what?? We don't know what caused the
numbers or the differences between them, or even if the first set of numbers are
typical. Please detail why you believe the entire reduction to be due to the
lights
2. It is not
always the case that increasing traffic increases risk. Once traffic reaches
saturation, and reduces to a crawl, (as it often does at the Fleetbridge
with lights) few significant accidents will happen at all. But creating
artificial traffic jams would not be my way to reduce accidents, and what about
the possibility that you've just moved them to somewhere else?. We cannot assume
that Tescos will have increased the accident rate. In any case, what was
different to the 3 years full time after installation, and the years from 2001
when full time was restored, in terms of contribution of safety of the lights?
3. There is no
error in the blue curve I added to the chart - it only illustrates what "might
have happened". It has to be a curve, not a straight line, as it cannot cross
the X axis. I did not offer it as proof of anything.
4. I drive
through this roundabout at least once a day, and using different entries /
exits. While the lights were off, sometimes a short stop before entering was all
I noticed, and then I could flow around it and exit without stopping. Now I have
to frequently stop in a queue to enter, then stop (quite often twice) as I go
round. This seems to be the experience of the vast majority as seen in online
news items. There is no danger if you look and give way if necessary. As I say,
if a driver needs lights to help them to do this, they're more than likely just
going to have that accident up the road instead. We can't idiot proof every inch
of every road (but we should be doing something about idiot drivers) .
5. I'm sure
speeds did increase without the lights, that is what everyone is saying, and why
they don't want the lights! But if you think that speed is a
problem here,
this needs to be dealt with by limits / enforcements, not by making traffic stop
every few yards. And in any case, a speed up to the limit (30) is entirely
reasonable. It's a very safe junction with ped / cycle underpasses and barriers.
It wouldn't surprise me if there was an accident due to someone forgetting the
lights had been deactivated, as you say, it needs time for changes to
settle. I'd like to see the evidence that traffic reduced and by how much.
Anyone not confident enough to use this roundabout without lights is a danger
and would benefit far more from some (re)training than the thousands sent on
courses for doing less than 40 on perfectly good, non residential dual
carriageways etc.
6. By "overall
experiment" I include the apparently significant increase that seemed to occur
when the lights were part time.
7. It is
interesting that the rates increased during part time operation. It would be
good to try to understand this, it might reveal something about what causes /
prevents accidents.
8. Who
determines if 3,4,5,6,7,8 accidents a year at such a junction is or is not safe
and how? It is impossible to imagine any traffic lights system that would
maintain a similar flow to that without lights, at a roundabout, which has
highly effective, efficient, natural prioritized flow.
As you see I am
not comforted and would welcome further responses on this. The main point I
would like to make is that the best money that can be spent would be on
improving driving standards which are simply appalling, and easily recognized,
and would bring benefits on ALL roads, junctions and roundabouts. This is what
is so unsatisfactory about our current road policing.
Sent: Wednesday,
December 07, 2011 3:59 PM
Subject: FW:
Fleetsbridge Roundabout Traffic lights - a proper analysis - traffic lights
probably INCREASED accidents -POOLE COUNCIL PLEASE RESPOND
Dear Mr Belchamber
Thank you for your emails dated 29 November and 1 December concerning the above.
Our aim is to respond within 10 working days of receipt of emails of this type
and this response is well within that target for your original email.
As discussed on 15th April 2011 you
offered to direct all emails which relate to road safety issues within Poole
directly to me. Please can you keep to this arrangement when using emails to
enable me to continue to assist you as I am unable to respond to further emails
which are not directed to me alone or do not relate to Poole.
In response to your specific points raised:
1.
Prior to the traffic lights being put in place at the junction there was
a significant accident record, which rose even higher when there was a trial
period of part-time operation in the late 1990’s.
I have carried out a statistical analysis of the 5-year period before and
after full-time signals were operated by checking All Purpose Poisson
Probabilities (Single factor Values) based on the long term average accident
rates for these two 5-year periods. It is possible to deduce from the figures
previously supplied to you that the reduction achieved is statistically
significant and that there is a high to very high confidence of a decrease below
the long term average following the switch to full time operation.
2.
The increases experienced around
1993-4 can be explained by the nearby Tesco store opening which generated
significant changes in traffic volumes into and out of the roundabout. The
signals were indeed installed to anticipate and manage these flows and to enable
better linking with the (then) new Tesco signals on Waterloo Road.
3.
Looking at the chart compiled by
you (presumably based on the data we supplied under your FOI request) it is
basically correct except for one noticeable error – you have not explained how
you have arrived at your light blue ‘curve’ line, but it is presumably simply a
version of the trend line as shown in my chart recreation below
(reproduced with the exception of serious accidents – there has only been one
since 2000 when the signals went full time).
I have also included in the
chart the equivalent annual fluctuations for Poole as a whole over the same
period (the additional green line on the chart). Basically a trendline is a only
a line made to fit for the data points given and is a best fit mathematical
equation for the chart. It should not be used to prove or disprove that
accidents would have dropped to this level anyway without signals operating.
4.
In situations such as those
encountered at this roundabout, signal priority will inevitably be given to the
major through routes. Therefore if you are entering the roundabout from one of
the minor roads, eg Fleets Lane, to exit onto Wimborne Road, then progression
around the roundabout will not be as continuous as if you were going from say
Waterloo Road to Wimborne Road. Having said this signals will benefit drivers
entering from arms of the roundabout with lower traffic flows which cannot be
safely be achieved by other means.
5.
Without signals in place during
the recent maintenance works speeds on the junction were increasing, with a
likelihood of increased casualty severity due to higher speeds. We have records
of at least one injury collision occurred here shortly after the signals were
switched off on the 31st October. In addition we have evidence that traffic
flows reduced significantly at the junction as drivers, particularly those less
confident, were taking alternative routes to avoid the area during the
roadworks. Just under 16,000 fewer vehicles a week were using the three main
arms of the junction – Waterloo Road, Upton Road and Wimborne Road. The current
changes being made to the operation of the signals will result in significant
savings in both energy and communications costs. We are moving over to Extra Low
Voltage and Broadband, as well as a more efficient method of signal operation at
off-peak times. Traffic volumes are now returning to normal.
6.
I do not agree with your
conclusion here because I believe your assumption is incorrect for the reasons
highlighted in 3 above. The current traffic control system is still bedding in
as the Siemens traffic signal engineers need time to fine tune the system based
on actual site observations at various times of the day and night. This is the
case with all signal junctions following “switch on”. If you use the roundabout
now I think you will find the situation improved compared with the period when
they were first switched back on again. I can of course understand the requests
for changes to the current system, but because of the potential safety
implications any decision should not be taken lightly. In this respect the
operation of the roundabout will be monitored and a report prepared for
consideration early in the New Year.
7.
The decision to trial part time
operation was something that was reached before I started in my current role,
but it is not something that is recommended by DfT, due to the very fact that an
increase in accidents is likely (as you have indeed observed). As I mentioned in
5 above a final and binding decision on operation will be made by Members in the
New Year.
8.
I would repeat my response to
your similar question raised at 4 above. Traffic signals will benefit drivers
entering from arms of very busy roundabouts with lower traffic flows which
cannot be safely be achieved by other means.
9 & 10. I refer to the response
given at 5. above.
Regards
Martin Baker
Senior Engineer (Road Safety and Urban Traffic Control)
T: 01202 262073 F: 01202 262091
E:
m.baker@poole.gov.uk
W:
www.boroughofpoole.com
Transportation Services, Borough of Poole, St Johns House, Serpentine Road,
Poole BH15 2DX
www.dorsetroadsafe.org.uk
From:
Ian Belchamber (gmail) [mailto:ianbelchamber@gmail.com]
Sent: 01 December 2011 08:20
Importance: High
Mr Baker, Mr
Tite, Poole Council,
I genuinely
hope that I'm wrong about this. The implications go far beyond just one
roundabout, if Poole have got this wrong, what about all of the other things it
has done, speed limit reductions, speed humps, other engineering "improvements".
What about it's funding of Dorset Road Safe? Is the logic Poole has used to do
these things equally as flawed? What about all the money wasted, the deaths not
reduced?
If Poole
Council believes it is right, why has there not been a swift and credible
response?
Please now
respond to this. If there is no adequate response I will of course be raising
complaints about this.
Regards, Ian
Belchamber
From:
Ian
Belchamber (gmail)
Sent: Tuesday,
November 29, 2011 1:42 PM
Subject: Re:
Fleetsbridge Roundabout Traffic lights - a proper analysis - traffic lights
probably INCREASED accidents
Sorry, point 1
should obviously be FROM 7 to 2.68, not BY 7 to 2.68. The remaining text is
unaffected. I have made some other minor improvements in the online version
http://www.dorsetspeed.org.uk/news/sog86.aspx
From:
Ian Belchamber (gmail)
Sent: Tuesday,
November 29, 2011 7:34 AM
Subject: Fleetsbridge
Roundabout Traffic lights - a proper analysis - traffic lights probably
INCREASED accidents
-
There is not
enough information to be certain about anything, other than one thing: to claim
that the lights have reduced accidents by 7 to 2.68 (which seems to be the ONLY
justification from Poole Council) is clearly complete nonsense:
-
Event rates are low, random variations high
-
Important factors such as traffic volumes / distributions, general trends,
changes in reporting, have not been considered
-
Other changes at the roundabout (road markings, effects on journeys by changes
at the many nearby large stores), etc have not been considered
-
However, if we
are to believe that the accident counts have been influenced predominantly by
the traffic lights, one thing (only) sticks out like a sore thumb: if you
include the few years of data PRIOR to the original installation of the lights,
the original installation (full time) in June 93 saw an INCREASE and when the
lights were switched to part time in Nov 95 accident counts increased again
significantly.
-
When the lights
were switched back to full time from Nov 2000, accidents did fall significantly
– but ONLY back to levels that might have been expected anyway if the lights has
never been there (as illustrated by the curve). And you can reduce accident
counts in one area by reducing flow, but those accidents will probably just
happen somewhere else.
-
The minimal
reductions seen could probably be achieved just by reducing flow by some other
means, reducing number of / widening lanes, etc, a bit of paint instead of
expensive traffic lights that cost drivers and the environment 24 hours a day.
This would not reduce flow at all when traffic is light and would limit flow,
vehicle density, make it easier to join leave the roundabout etc thus reducing
collisions when busy.
-
There are other
ways to reduce accidents: proper policing would be the most effective way. Even
if the lights have reduced accidents, it does not mean it is the most cost
effective and efficient way of doing it. All options should be properly
considered.
-
So a proper
analysis using all available data and looking at the wider picture can only
conclude that the overall “experiment” to install traffic lights on this
roundabout has caused more accidents than it has prevented, not to mention all
of the congestion, pollution, wasted time and fuel for thousands of drivers
every day.
-
Even despite
the uncertainties mentioned in 1 above, the accident increase during part time
operation looks undeniable. This surprises even me, perhaps there is an
explanation for this, maybe the presence of lights gives drivers a false sense
of security, without the control if they are switched off. Perhaps speed limits
/ humps / loads of road signs etc. have the same effect. Perhaps it is better to
allow drivers to think a bit more about their driving rather than blindly
following increasing numbers of rules and control that seem to have no reason
(but can of course generate revenue).
-
So,
reassuringly, a proper analysis supports common sense: it’s a roundabout, it
does not need traffic lights to improve safety OR help with traffic flow.
-
It is truly
worrying that our “experts” can get things so completely wrong, and draw the
conclusions it wants, to do what it wants, by such naïve and simplistic
interpretation of data. And these are the same “experts” presumably who are
deciding on new town bridges, road safety, speed limit reductions, other road
“improvements”, etc. No wonder we’re in such a mess. When are things going to
change in Poole / Bournemouth / Dorset?
-
The best course
of action now at Fleetsbridge is to simply remove the new traffic lights and
sell them for scrap, and to learn from the failure and to not repeat it. Please
can Poole Council do this or provide a proper scientific justification
for the lights.