Late at night, the traffic lights on Finchley Road at the junction of West
Heath Avenue often go red for a couple of seconds before reverting to green.
This is far too short for another traffic or pedestrian phase to have
occurred in between. It seems that sensors facing into the side roads detect
that there is no traffic there and that this causes the green phase for the
side roads to be skipped, but unfortunately the other lights have flipped to
red before this decision is made. Surely this must be a mistake rather than
design.

How are the phases of traffic lights controlled anyway? Is there a program
inside them written in some standard programming language, and someone has
put the IF statement in the wrong place? I would have guessed that the
danger of conflicting greens would have prevented the control sequence from
being written in a standard language and that something much less flexible
would have been used, which would have prevented the error I see at this
junction.

"Basil Jet" wrote in message
...

Late at night, the traffic lights on Finchley Road at the junction of West
Heath Avenue often go red for a couple of seconds before reverting to
green. This is far too short for another traffic or pedestrian phase to
have occurred in between. It seems that sensors facing into the side roads
detect that there is no traffic there and that this causes the green phase
for the side roads to be skipped, but unfortunately the other lights have
flipped to red before this decision is made. Surely this must be a mistake
rather than design.

It is quite common for lights to run though a short cycle when they have not
been triggered for some time. I presume the controller is running some sort
of error check sequence.

Colin Bignell

"Basil Jet" wrote in message
...

Late at night, the traffic lights on Finchley Road at the junction
of West Heath Avenue often go red for a couple of seconds before
reverting to green. This is far too short for another traffic or
pedestrian phase to have occurred in between. It seems that sensors
facing into the side roads detect that there is no traffic there and
that this causes the green phase for the side roads to be skipped,
but unfortunately the other lights have flipped to red before this
decision is made. Surely this must be a mistake rather than design.

It is quite common for lights to run though a short cycle when they
have not been triggered for some time. I presume the controller is
running some sort of error check sequence.

If it's common, where else does it happen? There is a pedestrian crossing in
Harrow which has phantom pedestrian phases, but apart from that, I've never
noticed it elsewhere.

"Basil Jet" wrote in message
...
"Basil Jet" wrote in message
...

Late at night, the traffic lights on Finchley Road at the junction
of West Heath Avenue often go red for a couple of seconds before
reverting to green. This is far too short for another traffic or
pedestrian phase to have occurred in between. It seems that sensors
facing into the side roads detect that there is no traffic there and
that this causes the green phase for the side roads to be skipped,
but unfortunately the other lights have flipped to red before this
decision is made. Surely this must be a mistake rather than design.

It is quite common for lights to run though a short cycle when they
have not been triggered for some time. I presume the controller is
running some sort of error check sequence.

If it's common, where else does it happen? There is a pedestrian crossing
in Harrow which has phantom pedestrian phases, but apart from that, I've
never noticed it elsewhere.

Phantom pedestrian phases are often cause by the local scrotii jamming a
match in the button....

On Jul 19, 9:02*am, "Ian" wrote:

"Basil Jet" wrote:

"Basil Jet" wrote:

Late at night, the traffic lights on Finchley Road at the junction
of West Heath Avenue often go red for a couple of seconds before
reverting to green. This is far too short for another traffic or
pedestrian phase to have occurred in between. It seems that sensors
facing into the side roads detect that there is no traffic there and
that this causes the green phase for the side roads to be skipped,
but unfortunately the other lights have flipped to red before this
decision is made. Surely this must be a mistake rather than design.

It is quite common for lights to run though a short cycle when they
have not been triggered for some time. I presume the controller is
running some sort of error check sequence.

If it's common, where else does it happen? There is a pedestrian crossing
in Harrow which has phantom pedestrian phases, but apart from that, I've
never noticed it elsewhere.

Phantom pedestrian phases are often cause by the local scrotii jamming a
match in the button....

Really? I've never come across that before.

"Ian" wrote in message
...

If it's common, where else does it happen? There is a pedestrian crossing
in Harrow which has phantom pedestrian phases, but apart from that, I've
never noticed it elsewhere.

Phantom pedestrian phases are often cause by the local scrotii jamming a
match in the button....

Hmm. No it should time out after a while and should send a constant input
alarm I think.

New crossings are button and beacon input - so you can push the button but
if nothing activates the sensor it should just 'wait' on calling the
pedestrian route till it gets an input then it will call it in the usual
way.

Junctions that go mad its often the Scoot loops broken/beacon, or something
in the chain broken and the controller is missing an input or another
action. Lights stuck on a phase usually send an alarm back after (I think
120 Sec on red)

Its also worth pointing out that a number (quite a lot) of junctions and
things are 'on-line' anyway feeding the Scoot data back to the GIS
application. The knock on of that is you can then 'control' the junction
(usually with defined configurations against time periods)

If you want to report the fault/problem then its TfL streets you want - and
they will either take the road junction details, or on the side of the
controller cab there is a number (XX/XXX) that's the site specific ID and
they will be quite happy with that too.

On Thu, 16 Jul 2009 00:23:13 +0100
"Basil Jet" wrote:
red before this decision is made. Surely this must be a mistake rather than
design.

Don't bet on it. perhaps it wasn't intentional to start with but when it was
realised it would slow traffic down they probably decided not to fix it.
The lights on the purley way in croydon still go red to let imaginary cars out
of empty shop car parks and industrial estates at 1am. Don't tell me thats
not done on purpose just to slow down traffic who might be trying to get a
move on on an empty dual carraigeway. You're not allowed to get anywhere
quickly in london.

How are the phases of traffic lights controlled anyway? Is there a program
inside them written in some standard programming language, and someone has

They probably used to run off an 4 or 8 bit microcontroller with the program
written in assembler, but these days it seems that you can't be a self
respecting embedded systems designer without specifying that the hardware must
be a PC running Windoze because the only thing you can code in is VB.

B2003

On 16 July, 11:39, wrote:
On Thu, 16 Jul 2009 00:23:13 +0100

"Basil Jet" wrote:
red before this decision is made. Surely this must be a mistake rather than
design.

Don't bet on it. perhaps it wasn't intentional to start with but when it was
realised it would slow traffic down they probably decided not to fix it.
The lights on the purley way in croydon still go red to let imaginary cars out
of empty shop car parks and industrial estates at 1am. Don't tell me thats
not done on purpose just to slow down traffic who might be trying to get a
move on on an empty dual carraigeway. You're not allowed to get anywhere
quickly in london.

How are the phases of traffic lights controlled anyway? Is there a program
inside them written in some standard programming language, and someone has

They probably used to run off an 4 or 8 bit microcontroller with the program
written in assembler, but these days it seems that you can't be a self
respecting embedded systems designer without specifying that the hardware must
be a PC running Windoze because the only thing you can code in is VB.

B2003

I should report it to TFL for London. I had a similar problem at a
junction a couple of years ago. it turned out the sensor for
detecting cars was broken so between 2200 and 0600 it would not let
any cars out of the side road. Had to pass through a red light
(carefully) with the permission of the police car who was waiting
behind me.

trainmanUK wrote:

TFL for London.

giggle

In article ,
wrote:
Don't bet on it. perhaps it wasn't intentional to start with but when it was
realised it would slow traffic down they probably decided not to fix it.
The lights on the purley way in croydon still go red to let imaginary cars out
of empty shop car parks and industrial estates at 1am. Don't tell me thats
not done on purpose just to slow down traffic who might be trying to get a
move on on an empty dual carraigeway. You're not allowed to get anywhere
quickly in london.

It's probably done on purpose on grounds of cost: lights with a
fixed (or timed) pattern will be cheaper to buy and maintain than
lights with sensors etc.

--
:wq