removing traffic engineering control – the awkward truth?

AN ALTERNATIVE FORM OF TRAFFIC MANAGEMENT?

www.tecmagazine.com TEC FEBRUARY 2011 73

Removing trafficengineering control – the awkward truth?

TRIALS AND TRIBULATIONSSome years ago, Ben Hamilton-Baillie, a UK expert in theideas of ‘shared space’ and simplified streetscape design,published an obituary to lowly traffic engineers, whowere to be put out to pasture (literally, as landscape gar-deners) by 2010 following the enlightened discovery thattraffic engineering was no longer required, because roadusers could quite happily optimise the use of road spaceand behave perfectly adequately without any need forthe modern traffic management paraphernalia; just sim-ple, aesthetic urban design. This was followed by anequally assaulting polemic from Martin Cassini, a TV pro-ducer and campaigner for traffic system reform, who as-serted that our need for establishing priority was com-pletely misplaced, and called for the removal of traffic sig-nals and all forms of junction control.

At about this time, the scheme in Ashford, Kent, wasbeing completed and certainly seemed to demonstratethat these ideas of shared space (that had been developedin Netherlands, Sweden and Germany, for example seeTEC Sep 2006)) could work in the UK. But it was going totake a great deal of effort to convince traffic engineersand traffic managers that the absence of formal control atbusy junctions might provide the best form of control.The only way would be through monitored trials at nu-merous sites to gather enough data and evidence to testthe Buchanan/Cassini hypothesis that:

At given junctions within a given road network, the removal(or absence) of automatic traffic signal control and standardpriority rules will have economic, social and environmentalbenefits, will improve road safety, vulnerable road-useramenity and traffic management functions.

Keith Firth

This paper analyses the results from the Cabstandjunction in Portishead, near Bristol, which failedfor a few hours in June 2009 and has since beenthe site of a grround-breaking experiment to

remove all junction controls. The Cabstand trial,still ongoing over a year later, has been followed bya further two trials in Bristol and the results fromthese trials are analysed. Although it is

acknowledged that further research is needednevertheless the trials demonstrated that despitetheir differences the junctions generally performedbetter without traffic signal or any formal control.

Advance warning at theCabstand junction inPortishead, near Bristol.

CONTACT DETAILSKeith Firthis Director -Traffic Engineeringwith Colin Buchanan.He can be contactedon +44 (0) 117 9170800 or by email [email protected]

Bristol[1[1].qxd 1/2/11 12:29 Page 73


74 TEC FEBRUARY 2011 www.LocalGov.co.uk

Although the idea of such trials was generally wel-comed by highway authorities we approached, and eachwould be very interested in any results gleaned, no au-thority was willing to undertake the risk. Until, that is, aparticularly unpopular set of MOVA controlled signals atthe Cabstand staggered crossroad junction in Portishead,near Bristol, failed for a few hours in June 2009. The re-sulting disappearance of queues and delays spurred thelocal press to run a front page story on how well the junc-tion seemed to perform without controls. After contact-ing the local councillor, David Pasley, and further discus-sions with North Somerset Council (NSC) engineers(Frank Cox and Ian Wilson), it was decided that thiswould be an ideal site for the first of these potentiallyrisky, but truly groundbreaking experiments.

NSC, with Colin Buchanan and Martin Cassini(CB/MC), worked on the trial and monitoring methodol-ogy, advised stakeholders and emergency services, carriedout a safety audit and risk assessment of the proposed ex-periment and devised an advance warning signing strat-egy. The trial date was set for September 2009 and wouldrun for up to four weeks, depending on the outcome atthe end of the first day. Monitoring of typical, controlledbehaviour at the junction started a week before the bigswitch-off.

Within hours of hooding the signals, things were look-ing bleak for the traffic engineering fraternity. Up to 2000vehicles per hour sailed through the junction with little,if any, delay and queues disappeared on all the ap-proaches. Drivers were courteous to each other, a goodproportion slowed to allow pedestrians to cross, and roadusers interviewed a few days before the trial who had saidit would be chaos, now reported that they were preparedto have a three-course millinery delight.

The signals at Cabstand have now been out of actionfor over a year. There have been just two reported slight(driver/passenger) injury accidents following minor colli-sions, as well as two damage-only incidents that we knowof, one of which occurred (most unfortunately) when theBBC were filming at the roadside for a report on the up-coming Bristol Trials. Zebra crossings have been installedat two crossing points to provide formal crossing facilitieson a key desire line, but other than that the trial demon-strated that removing all forms of conventional junctioncontrol resulted in less traffic congestion, fewer delaysand queues, and greater capacity, with little impact onpedestrian amenity. There are now plans to re-think thedesign of the space entirely to facilitate a more simplifiedstreetscape.

This led to a campaign by the local press to switch offtraffic signals across Bristol. With the immediate backingfrom the Executive Member for Transport at the time, LibDem Councillor Dr Jon Rogers, Bristol City Council(BCC) engineers (Terry Bullock, John Laite and AdamCrowther) were confronted with the task of identifyingsuitable trial sites. From a list of nine potential sites, twowere chosen that were considered not to be too risky, butat the same time would give a good indication of behav-iour and junction performance. These were Union Street/Broadmead/ Nelson Street and Broad Quay/ Marsh Street,both in the centre of the City. These were junctions thatdid not experience significant congestion problems, andso it would be interesting to see if lack of control couldperform better than typical, good quality vehicle andpedestrian actuation systems.

BCC and CB/MC organised stakeholder and road-usergroup consultation events, developed the trial and moni-toring methodology (to include a road-user satisfaction

survey) and a slightly different signing strategy to thatused at Cabstand. BCC were keen not to provide guid-ance on behaviour, but were interested in how road usersresponded to simple warning signs. The trials were heldin March 2010, with monitoring occurring a week beforeand a week after switch-off at each site. Unlike Cabstand,it was decided that the signals would be switched back onat the end of the trial period regardless of results, withany decisions on the implications being deferred untilafter CB submitted the monitoring reports.

Once again, we found ourselves writing a report thatconcluded that the disabling of all junction controls re-sulted in improvements to traffic capacity and reductionsin journey time for both vehicles and pedestrian alike,and hence a reduction in queues and delays. The road-user satisfaction surveys showed, however, that a highproportion of pedestrians did not feel as safe without con-trol and would prefer green man crossing facilities, eventhough most acknowledged there was less delay. Indeed,the idea of the lack of any formal control was particularlyopposed by local visually impaired groups and some dis-abled pedestrians. Nevertheless, during the weeks of thetrials there were no incidents or accidents.

The success of the trial in showing journey time bene-fits prompted BCC officers to recommend that the Execu-tive progress with the evaluation of additional sites acrossBristol.

TRIAL BY JURYThe trials clearly had to be developed with public safety atthe forefront of any strategy and methodology. Undernormal conditions of signal failure, Traffic Signs Manual/legislation requires posting of ‘lights out’ signing, whichis sufficient to advise motorists of the lack of control andthe need to proceed with caution. The trials, however,would need far greater consideration to the risk of acci-dents and litigation and the avoidance of any confusionby users of the junctions.

There was undoubtedly a risk of accidents, yet withoutany precedent upon which to base the experiment therewas no way of evaluating the risk other than engineeringjudgement. In any case, this risk had been accepted bythe respective Executive Members of the councils whenthe bold decision was made to proceed with the trials. Lit-igation against councils for embarking on experiments ofthis nature, or indeed for unconventional or even con-ventional traffic engineering and management solutionsis rarely, if at all, successful. It was evident that with ap-propriate planning and advance warning of the trials,there was no risk of successful litigation in the event of anaccident or incident.

The trial methodology developed by CB and the re-spective councils was scrutinised under standard EqualityImpact Assessment (EquIA) and Stage 1 Safety Audit. Inboth the Cabstand and Bristol trials, slight modificationsto the temporary highway arrangements were recom-mended by the Safety Audit, including the introductionof a 20mph zone around the Cabstand junction. Some ofthese proved somewhat detrimental and influenced thetrial results to the extent that, at Cabstand, the measuresto prevent pedestrians from using a particular crossingpoint were removed shortly after the trial commenced.

The signing strategy was considered fundamental tothe outcomes of the trial. Signing that was too prescrip-tive would not provide road users with the freedom to be-have in the altruistic manner that might be expectedunder a truly shared space environment. Ambiguoussigning might, however, introduce undesirable risk. In

Bristol[1[1].qxd 1/2/11 12:30 Page 74

both cases, advance warning of the signal switch-off trialswas provided on all approaches to the junctions, andmedia coverage ensured that a good proportion of roadusers were aware of the experiment in advance. At Cab-stand, drivers were advised to give way to pedestrians andthat there is no vehicle priority. At the Bristol sites, allusers were simply advised to proceed with care.

Wider public consultation on the value or methodol-ogy of the trial was not necessary, as the decision to goahead with the experiments was made at Executive levelwithin the councils, however it was considered importantto advise a range of road-user groups of the trials and ex-pected conditions. The greatest objections were, not sur-prisingly, made by vulnerable pedestrian groups, particu-larly those representing the blind and visually impaired.A major concern was the loss of formal crossing facilitiesintroduced for the purposes of social inclusion, nowbeing removed without any physical mitigation measureto replace them. The idea of inclusion through a changein social behaviour was not considered reasonable.

TRIAL BLAZING

The traffic signal control at Cabstand was introduced in 2005as part of a major public realm modification scheme associ-ated with a large residential development in the town. Thesignal control was introduced in order to facilitate formalpedestrian crossings and to allow for potential traffic growth.The long cycle time and complex staging arrangements,however, gave rise to long delays and the junction was oper-ating close to capacity during the peak hours, resulting inlong queues particularly back up the High Street approach. Apedestrian crossing demand of between 200-300 movementsper hour comprises schoolchilden, shoppers and commuters.Traffic demand was around 1500pcu/hr during the peakhours, with very few buses. A parallel residential street to thewest, Slade Road, was used as a rat-run by local traffic and soATC loops were installed to monitor any changes resultingfrom the trial. Two-way vehicle demand on this route wasaround 2000veh/hr during the peak hours.

Immediately following signal switch-off, vehicle delays



0

200

400

600

800

1000

1200

1400

1600

1800

2000

Veh

icle

dem

and

(PC

U/h

r)

Station Road Wyndham Way High Street Cabstand Total

Approach

Cabstand vehicle demand

Before switch-off AM

After switch-off (Week 1) AM


Before switch-off PM

After switch-off (Week 1) PM


0

50

100

150

200

250

300

Ped

estr

ian

cro

ssin

g m

ove

men

ts

(ped

s/h

r)

Station

Road

Wyndham

Way

High Street Cabstand Mid-junction

stagger

Total

Crossing location

Cabstand pedestrian demand

Before switch-off AM



Before switch-off PM



0

20

40

60

80

100

120

140

Veh

icle

qu

eue

len

gth

(P

CU

)

Station Road Wyndham

Way

High Street Cabstand Total

Approach

Cabstand vehicle queue length

Before switch-off AM Mean

Before switch-off AM Maximum

After switch-off (Week 1) AM Mean

After switch-off (Week 1) AM Maxiumum


After switch-off (Week 3) AM Maximum

Before switch-off PM Mean

Before switch-off PM Maximum

After switch-off (Week 1) PM Mean

After switch-off (Week 1) PM Maximum



00:00

02:53

05:46

08:38

11:31

14:24

17:17

20:10

23:02

25:55

28:48

Jou

rney

tim

e (m

in:s

ec)

Station Road WyndhamWay

High Street Cabstand Total

All journey start points

Cabstand vehicle journey time

Before switch-off AM Mean

Before switch-off AM Maximum


After switch-off (Week 1) AM Maxiumum


After switch-off (Week 3) AM Maximum

Before switch-off PM Mean

Before switch-off PM Maximum





SiteCabstand, Portishead

Street (Arm)A369 Wyndham Way (2-lane dual carriageway)/ B3124 HighStreet (south)/ Cabstand (west)/ Station Road (north)

Signal controlFull time MOVA, 6 traffic phases, 6 pedestrian/cyclist phases(including 2 staggered crossings)

Peak hour mean cycle time 125sec, maximum cycle time 160sec

Trial datesPre-trial surveys week commencing 7 September 2009Signal switch-off for 4 weeks commencing 14 September 2009

Site layout

Bristol[1[1].qxd 1/2/11 12:31 Page 75



and queues reduced by 50%. As a consequence, traffic de-mand through the junction grew by 20% to over 2000pcu/hrduring the peaks, with monitoring of the residential streetsshowing a compensating reduction in rat-running. Althoughthis led to a slight increase in journey times by Week 3, delayswere still broadly half the pre-trial values. Pedestrian demandfell during Week 3 due to poor weather conditions, so datafrom Week 1 was used to compare crossing times and delays.If pedestrians waited for the green man invitation, averagecrossing times would be expected to be around a minute,however surveys showed the mean crossing time to bearound 20 seconds, indicating that pedestrians rarely usedthe formal facilities provided. Following switch-off, meancrossing times were very similar to pre-trial values, indicatingthat general behaviour was hardly affected and pedestrianswere content to cross in gaps appearing, or that were pro-vided, in traffic streams. What was evident, however, was thatthe maximum crossing times reduced in most cases, givingaverage reductions of at least 20%.

Interestingly, post-trial traffic capacity analysis using AR-CADY showed that the introduction of mini-roundaboutmarkings would be likely to create internal queue storageproblems, something that has not occurred under uncon-trolled conditions.

The traffic signal control at Union Street mainly pro-vides controlled crossing facilities for the high volume ofpedestrians moving between Nelson Street and the Broad-mead pedestrianised shopping street, yet also provides abus gate facility and manages northbound vehicular traf-fic. Traffic queues from the downstream junction canblock back through the junction, although generally thejunction performs well. Vehicular demand is between500-600pcu/hr, including around 100 buses, and thereare usually over 3000 pedestrian crossing movementsduring the peaks. The number of cyclists is generally low,at around 20-30 per hour.

During the period of signal switch-off, vehicle andpedestrian demand was generally higher, although thiswas not felt to be as a consequence of the trial, rather pat-terns in shopping behaviour. Despite this increase, meanvehicle queues and delays reduced by 30%. Pedestrian be-

0

100

200

300

400

500

600

700

800

Veh

icle

dem

and

(PC

U/h

r)

UnionStreet NelsonStreet Total

Approach

Union Street vehicle demand

Before switch-off weekdaypeak

After switch-off weekdaypeak

Before switch-off Saturdaypeak

After switch-off Saturdaypeak

0

500

1000

1500

2000

2500

3000

3500

4000

Ped

estr

ian

cro

ssin

g

mo

vem

ents

(ped

s/h

r)

Union Street Nelson Street Total

Crossing location

Union Street pedestrian demand

Before switch-off weekday peak

After switch-off weekday peak

Before switch-off Saturday peakAfter switch-off Saturday peak

0

5

10

15

20

25

Veh

icle

qu

eue

len

gth

(PC

U)

Union Street Nelson Street

Approach

Union Street vehicle queue length

Before switch-off weekdaypeak Mean maxBefore switch-off weekdaypeak MaximumAfter switch-off weekdaypeak Mean maxAfter switch-off weekdaypeak MaxiumumBefore switch-off Saturdaypeak Mean maxBefore switch-off Saturdaypeak MaximumAfter switch-off Saturdaypeak Mean maxAfter switch-off Saturdaypeak Maximum

0

20

40

60

80

100

120

140

Veh

icle

jou

rney

tim

e (s

ec)

Union Street Nelson Street Total


Union Street vehicle journey time

Before switch-off weekday peak Mean

Before switch-off weekday peak Maximum

After switch-off weekday peak Mean

After switch-off weekday peak Maxiumum

Before switch-off Saturday peak Mean

Before switch-off Saturday peak Maximum

After switch-off Saturday peak Mean

After switch-off Saturday peak Maximum

0

20

40

60

80

100

120

Ped

estr

ain

cro

ssin

g m

ove

men

ts

(pe

ds

/hr)

Union Street

eastbound

Union Street

westbound

Nelson Street

northbound

Nelson Street

southbound

Total

Crossing location

Pedestrian crossing time

Before switch-off weekday peak Mean

Before switch-off weekday peak Maximum

After switch-off weekday peak Mean

After switch-off weekday peak Maximum

Before switch-off Saturday peak Mean

Before switch-off Saturday peak Maxiumum

After switch-off Saturday peak Mean

After switch-off Saturday peak Maximum

Site

Union Street, Bristol

Union Street (one-way northbound)/ Nelson Street (one-wayeastbound)/ Broadmead (pedestrianised)

Non-UTC VA and pedestrian actuation, 2 traffic phases, 2pedestrian phases

Peak hour typical cycle time 40sec

Pre-trial surveys week commencing 28 February 2010

Signal switch-off for 1 week commencing 7 March 2010

Street (Arm)

Signal control

Trial dates

Site layout

Bristol[1[1].qxd 1/2/11 12:31 Page 76

haviour varied between the weekday and Saturday peakperiods, with more green man compliance on a Saturday.This meant that the reduction in both mean and maxi-mum crossing times following switch-off was moremarked on the Saturday, at around 30%, yet there werebenefits during the rest of the week.

Road-user satisfaction surveys showed that two thirdsof all those surveyed (mostly pedestrians who had nottravelled through the junction by car) believed the junc-tion to be safer and easier to use under signal control, and75% of all respondents would prefer the signals to beswitched back on, yet interestingly only half recognisedthat there were fewer delays without signal control.

The traffic signal control at Broad Quay was introduced



SiteBroad Quay, Bristol

Street (Arm)Marsh Street (north)/ King Street (minor access road)/ PrinceStreet (south)/ Broad Quay (bus/cycle only to northwest)

Signal controlNon-UTC VA and parallel stage stream exit crossing, 6 trafficphases, 4 pedestrian phases (including 1 staggered crossing)

Peak hour typical cycle time 50sec

Trial datesPre-trial surveys week commencing 14 March 2010

Signal switch-off for 1 week commencing 21 March 2010

Site layout

Union Street response on feeling of safety

Safer w ith traf f ic s ignal control

Safer w ithout traf f ic signal control

No dif ference

Don'tknow

Union Street response on ease of use

Easier w ith traf f ic signal control

Easier w ithout traf f ic signal control

No dif ference

Don'tknow

Union Street response on delay

Quicker w ith traf f ic s ignal control

Quicker w ithout traf f ic signal control

No dif ference

Don'tknow

0%

10%

20%

30%

40%

50%

60%

70%

80%

YES NO Don’tknow YES NO Don’tknow

snairtsedePsrevirD

Union Street - Should traffic signals be switched off?

0

100

200

300

400

500

600

700

Veh

icle

dem

and

(P

CU

/hr)

Marsh Street Prince Street Broad Quay Total

Location

Broad Quay vehicle demand

Before switch-off AM peak

After switch-off AM peak

Before switch-off PM peak

After switch-off PM peak

Hoodingsignals inpreparationfor theUnionStreet trial.

Bristol[1[1].qxd 1/2/11 12:32 Page 77



in 2006 and incorporated bus lanes, a bus gate and all-round controlled pedestrian crossing facilities. Traffic cansometimes queue from the downstream junction to thenorth and block back through the junction, althoughgenerally the junction performs well. Vehicular demandis around 600pcu/hr (including around 100 buses) andthere are around 300 cyclists and 1000 pedestrian cross-ing movements during the peaks. The junction is locatedon a popular through-route for pedestrians and cycliststravelling between the Temple Meads station and quay-side areas of Bristol. A lot of pedestrians tend to use gapsin traffic created by the signals to cross all around thejunction, and not just at the formal crossing points.

Following signal switch-off, vehicle, pedestrian and cy-clist demand were unaltered. The results show that inmost cases, and therefore overall, mean and maximumjourney times reduced by around 30%, with mean queuelengths reducing by 40% and yet, rather interestingly,maximum queue lengths were not as greatly improved.Mean pedestrian crossing times reduced by a few secondsand thus, overall, by around 10% and the maximumcrossing times were reduced by at least 20%.

Road-user satisfaction surveys showed that most feltthe junction to be safer, easier to use and quicker to passthrough without signal control, and some 70% wouldprefer to keep the signals switched off. Of those thatwould prefer signal control, all acknowledged that it wasquicker, or certainly no slower, without controls.

These results were almost the complete opposite of theviews from respondents at Union Street, yet the junctionsare located less than 2km apart. It was felt that this wasdue to the difference in the type of pedestrian at each ofthe sites. At Union Street, those surveyed were part ofgroups or families with children out shopping on a Satur-day, not necessarily familiar with the junction. At BroadQuay, pedestrians tended to be alone or as part of a pairon a regular and familiar commuting trip.

TRIAL CALMINGAverage vehicle speeds were monitored at each of thesites. At Cabstand, a 20mph zone had been introduced as

0

100

200

300

400

500

600

700

800

900

1000

Ped

estr

ian

cro

ssin

gm

ove

men

ts(p

ed/h

r)


Crossing location

Broad Quay pedestrian demand

Beforeswitch-offAMpeak

After switch-off AMpeak

Beforeswitch-offPMpeak

After switch-offPMpeak

0

2

4

6

8

10

12

14

16

18

20

Ve

hic

leq

ue

ue

len

gth

(PC

U)


Approach

Broad Quay vehicle queue length

Before switch-off AM peak Mean max

Before switch-off AM peak Maximum

After switch-off AM peak Mean max

After switch-off AM peak Maxiumum

Before switch-off PM peak Mean max

Before switch-off PM peak Maximum

After switch-off PM peak Mean max

After switch-off PM peak Maximum

0

20

40

60

80

100

120

140

160

180

Jou

rney

tim

e(s

ec)



Broad Quay vehicle journey time

Before switch-off AM peak Mean


After switch-off AM peak Mean

After switch-off AM peak Maxiumum

Before switch-off PM peak Mean

Before switch-off PM peak Maximum

After switch-off PM peak Mean


0

20

40

60

80

100

120

140

160

180

Ped

estr

ian

cro

ssin

g m

ove

men

ts

(pe

ds

/hr)

MarshStreet

eastbound

MarshStreet

westbound

PrinceStreet

eastbound

PrinceStreet

westbound

Broad Quaynorthbound

Broad Quaysouthbound

Total

Crossing location

Broad Quay pedestrian crossing time

Before switch-off AM peak Mean


After switch-off AM peak Mean

After switch-off AM peak Maximum

Before switch-off PM peak Mean

Before switch-off PM peak Maxiumum

After switch-off PM peak Mean


Broad Quay response on feeling of safety

Safer w ith traf f ic signal control

Safer w ithout traf f ic s ignal control

No dif ference

Don't know

Bristol[1[1].qxd 1/2/11 12:33 Page 78

a consequence of Safety Audit recommendations, yetspeeds may have remained at around 15mph, it was felt,due to the geometric constraints of the staggered junc-tion arrangement. At Union Street, speeds stayed ataround 25mph, which was perhaps higher than might beexpected, but probably due to the downhill approach tothe junction. At Broad Quay, speeds reduced slightly toless than 20mph, perhaps indicating that the signalswitch-off resulted in a traffic calming effect.

CONCLUSIONSWith data from only three trial sites where the effects ofremoving traffic control regulations, in these cases signalcontrol, have been monitored (despite the significantlevel of scrutiny), it is not really possible to provide proof,or otherwise, of the Buchanan/Cassini hypothesis. It isclear that further research at a much greater number ofjunctions and network arrangements and over a longerperiod of time will be required before any satisfactoryconclusions can be drawn. The trials have not been in op-eration for long enough to understand impact on roadsafety, and the issue of appropriate crossing facilities forvulnerable road users needs to be addressed rationally.

Nevertheless, the trials have all demonstrated that de-spite their differences the junctions generally performedbetter without traffic signal, or indeed any, formal con-trol. Vehicle delays and pedestrian crossing times gener-ally reduced under the shared space arrangement, tovarying degrees depending on how well the junction per-formed previously under formal controls; where the junc-tion is operating efficiently, only small improvementswere found. Yet this is perhaps the most important con-clusion, that removing or not providing formal controlsat busy, urban junctions seems to offer a legitimate formof traffic management, that may not be any worse thanconventional priority or signal controlled methods, andindeed may show significant benefits. Further benefitsmight also be achieved through a greater public under-standing and acceptance of uncontrolled, shared spaceprinciples.

CB has used the wealth of data collected during the tri-

als to develop innovative micro-simulation modellingtechniques to forecast how other, even busier junctionsmight perform, and this has shown some limitations towhere we believe shared space principles might be ap-plied. Yet this research has demonstrated that there is afar greater opportunity to develop designs and trafficmanagement strategies with emphasis on the characteris-tics of the ‘place’, rather than traffic management tech-nology, than might previously have been thought.



Broad Quay response on ease of use


Easier w ithout traf f ic s ignal control

No dif ference

Don'tknow

Broad Quay response on delay


Easier w ithout traf f ic s ignal control

No dif ference

Don'tknow

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

YES NO Don’t know YES NO Don’t know YES NO Don’t know

snairtsedePstsilcyCsrevirD

Broad Quay - Should traffic signals be switched off?

0

5

10

15

20

25

30

Ave

rag

esp

eed

(mp

h)

Cabstand Union Street Broad Quay

Trial site

Average vehicle speeds through junctions

Average speed before switch-off

Average speed after switch-off

The Broad Quayjunction duringswitch-off.

Bristol[1[1].qxd 1/2/11 12:33 9

removing traffic engineering control – the awkward truth?

Documents