Traffic Lights and Multi-Aspect Signals

View previous topic View next topic Go down

Traffic Lights and Multi-Aspect Signals

Post by johnsinden on Tue Jun 19, 2012 10:56 pm

Had a debate in the office today, but couldn't resolve it. Almost everywhere in the world traffic lights follow the same pattern, i.e. red is at the top, amber in the middle and green at the bottom. Yet on the railways which use colour light signals, the sequence is generally the reverse, i.e red is at the bottom, yellow in the middle, and green traditionally at the top of the cluster. Does anyone know why this is? Answers please on a postcard......................................!

John
avatar
johnsinden

Posts : 210
Join date : 2011-09-19
Age : 64
Location : Southampton, UK

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by Quork on Tue Jun 19, 2012 11:17 pm

That's true with UK and countries formerly occupied and colonized by the UK. In other countries, it's different. Remember how I once narrated the history of german Hl system? It's based on an international agreement of the east block-pendant to the UIC, virtually an international system. You can find it e.g. in former GDR, in Poland or in Hungary. I'll leave pure distant signals out as they are not interesting (only green and yellow), let's look on combined main+distant signals: The aspect is two-parted. The upper half is the distant signal, the lower part the main signal, so to say. When the signal is at danger, there is no distant signal information (what would it be for?); thus red is in the middle. Below you have green and yellow (in german Hl system one next to the other, horizontally, in Poland or Hungary one over the other) for the main signal function ("here"), and above also yellow and green for the distant signal function ("on the next one").

Other example; german H/V light system, which is the night (light) signals of the classic H/V semaphores, thus the layout of a main signal is:

G
R R

Y

So again, another sequence. Next signal system; Ks. Three lights in a triangle; red at the top, green at the left, yellow at the right.

To add obfuscation to confusion; these are only the usual layouts, but that's not written in stone and often enough different depending on local parametres (curve, tunnel, bridge, etc. etc.); the only thing written in stone is the aspect, e.g. "green over yellow" for Hp2 or "red" (or "red next to red", there was a difference between Hp0 with one and Hp00 with two red lights, but that's obsolete since german reunion, so you see both versions) for Hp0.

Why is that? Well, in street traffic it is important the red light is seen as soon as possible, that's why it's upmost, while the others are below. On the railway however, that's of no concern. You don't have to see the signal aspect from as far as possible because of distant signalling, thus visibility distance is of no importance. This way, the sequence of colours is dictated by historical (H/V) and technical parametres and varies from country to country and from signal system to signal system.

On a side note; leaving semaphores aside, it seems Germany is the only country with several signalling systems, am I right? For all I see and read in the OpenBVE world, it seems the UK, France, Italy, Hungary, Ireland, Japan etc. pp. all have only one light signalling system; and for Poland I know there is virtually only that international east block one, with very few old german H/V light signals left on some heritage lines in west-most parts of the country.
avatar
Quork

Posts : 1087
Join date : 2012-05-05
Age : 26
Location : Hofheim a.T., Hessen (Hesse), European Union

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by graymac on Wed Jun 20, 2012 1:02 am

Cos the red aspect is most directly in the drivers sight line, nearest eye level as the train approaches the signal head. To have the other aspects lower would detract from longer distance viewability
avatar
graymac

Posts : 1741
Join date : 2011-08-28
Location : Co Mayo, Eire

http://www.celtictrainsim.com

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by mrknowitall on Wed Jun 20, 2012 1:18 am

Simple really , guy was creating on paper, woman gave him greif during the Process, now everythings upside down Smile

Does that mean if a train driver is driving his car and jumps the light and gets pulled, could he get away with "LOS" been an issue? Smile or should i know that? Smile

Getting tired now its early moring, ive only read 6 posts by quork but by heck, my eyes are stinging cause he types so much Smile
avatar
mrknowitall

Posts : 711
Join date : 2011-07-09
Age : 25

http://www.bve-terminus.org/pioneertrains/

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by johnsinden on Wed Jun 20, 2012 6:59 pm

Yes, see what you mean. @Quork - thank you for your dissertation. Just a slight but significant correction - this has nothing to do with colonization or occupation, many countries throughout the world approached Great Britain for assistance with railway construction and more often than not Robert Stephenson got the job, which explains why the trains in Italy and Japan, for example, travel on the left.

Interesting that we have differing views (excuse the pun). Is it to assist visibility, or could it even be to avoid confusion between road and rail signals, particularly if road and rail run parallel to each other. Good Pub Quiz, anyway.

Cheers,

John
avatar
johnsinden

Posts : 210
Join date : 2011-09-19
Age : 64
Location : Southampton, UK

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by James on Wed Jun 20, 2012 8:28 pm

A couple of clarifications about UK practice - it isn't that the red is at the bottom. It's that the red is closest to the Driver's eye-line (as graymac says). Most signals are above the Driver's eye-line, so this rule does indeed put the red at the bottom. However, if the signal is located below the Driver's eye-line (e.g. in tunnels signals are often located on the ground) then the red is normally at the top - as that is then closest to the Driver's eye-line.

There are exceptions. under station canopies, for example, the lamps may be arranged in a horizontal, rather than a vertical line, because there isn't the space to arrange them horizontally. In that case (assuming the signal was to the left of the running line) the red aspect would be at the right-hand end of the signal (again - closest to the Driver's eye-line). I think there's an example of such a signal on Anthony Bowden's Cross City South, at Birmingham New Street.

Quork - in answer to your question about different signalling systems within a country: the UK has some other systems such as RETB (Radio Electronic Token Block) where the only lineside equipment are fixed instruction signs (most just say 'Stop') and the Driver communicates with the signaller by radio. It also has a number of different ways of displaying signal aspects. For example, there are single lens ('searchlight') signals which show all 3 aspects through one lens, rather than having 3 separate lamps. The limitation of these, of course, is that they can't display a double yellow aspect, so if you want 4 aspect signalling you have to have an additional lens. Just when we thought the concept was dead, along came LED signals - and now, once again, whatever the colour of the signal it is emitted from the same point (difficult to describe, just google it for some images!)

Although you say 'leaving semaphores aside' it's important to note that in the UK there are two distinctly different ways of signalling, and both can be done using colour light signals - so as a Driver you must know under which system the line you are driving is operated. Most lines now are under the most simple form of signalling (seen on most BVE routes) which simply states that the signal sequence will be Green, Double Yellow, Single Yellow, Red for 4-aspect signalling, or Green, Single Yellow, Red for 3-aspect signalling. However, some lines are still signalled according to 'absolute block' principles, and suddenly a signal aspect can mean something different...

With absolute block a single distant signal can provide advance warning for a number of stop signals. What this means for the Driver is that if you pass a single yellow, and then a green, the next signal may still be red. You might even pass a single yellow, then two greens, and still find the next signal at red. As a Driver your route knowledge tells you how many 'stop' signals are protected by the distant signal, and having passed the distant signal at yellow you must be prepared to stop at any of the stop signals.

The UK remains an interesting location for the signalling enthusiast, with plenty of weird and wonderful variations for those prepared to search.

James

Posts : 61
Join date : 2011-08-22

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by Quork on Wed Jun 20, 2012 9:38 pm

Yeah it isn't always directly connected to colonization and occupation, but mostly it is countries which were colonized by the british crown and thus traditionally are seeking help in many matters there. (=> Commonwealth) And still, some, as formerly occupied southern Ireland or colonized India, had their railways built back then by the British and thus simply kept signals and rules, of course diverging since their regaining of independency.

What you described is still the same signalling system. It is in technically different signals, but still, the aspects look the same and mean the same, with that absolute blocking thing being a special case.

For example, all these are H/V light signals, although looking way different:


source: BahnBilder.de
The oldest type still in use


source: drehscheibe-foren.de / drehscheibe-online.de
A newer type


source: drehscheibe-foren.de / drehscheibe-online.de
One of the newest types


source: stellwerke.de
Old type again


source: stellwerke.de
former GDR variant (from before Hl system, only few were constructed)

See how they look different? Also, the position (whether above or below drivers eyeline) doesn't affect the sequence of lights - while other local parametres do.


Own pic
Intermediary signal R 154 of Stuttgart main station. Signal itself is the 1964 (I think) form, only lowered by four metres for better visibility.


Own pic
Exit signal N 6 of Rastatt main station. I only have this night photo, so you don't see well, but it's clearly visible it has been stuffend below the roof - the signal post plate (white-red-white) isn't on the signal post but left of it, and the subsidiary signal screen is on the left, above the Zp9 screen, instead of below the main signal screen. (Zp9 btw is the signal the train supervisor at stop stations (mostly the chief guard) gives to the driver when the train is ready to depart; it's the white, green-bordered signalling disc at day, a green hand lamp at night or, as here, a stationary signal showing a green ring activated with a key switch on the platform)


Another system is Hl:


source: sh1.org
The classical form with nearly all optics there (only the green or the yellow light bar is missing)


source: BahnBilder.de
A somewhat more compact form


source: bastim77.beepworld.de
Even compacter


source: drehscheibe-foren.de / drehscheibe-online.de
Special case


source: drehscheibe-foren.de / drehscheibe-online.de
EZMG form



The same goes for the other signalling systems. There is a vast amount of variations within each signalling system, but the system is the same. And for me as a driver it is not interesting in any way what the signal looks like. Triangular, on eight legs, with arms, a silly face and a green over a yellow, that's okay with me; the aspect is non-ambiguously Hp2. And whether it's on a five and a half metres signal post with a signal screen one metre and a half high or if it is the size of a banana box and standing on the ground, that's all the same with me as well (interesting enough, in difference to the Netherlands; a low-standing signal showing green means "proceed with 40km/h" while a high-standing signal showing green means "clear" there)



The other thing you mentioned is similarly known in Germany as well, it is called Zugleitbetrieb (literally: Train managing operation). It might be signal-assisted or not, depends on the route. Instead of complicated technical measures for protecting the trains everything is based on personal communication. The train manager has a linear depiction of the route and colour crayons. He draws a line on the map from the position of the train to a defined point, then he dictates a written permission to proceed till this point to the driver. After this, the driver proceeds to the specified point and requests further permission again.
avatar
Quork

Posts : 1087
Join date : 2012-05-05
Age : 26
Location : Hofheim a.T., Hessen (Hesse), European Union

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by James on Sat Jun 23, 2012 9:01 am

Thanks for all the information Quork - very interesting (and I like the idea of the crayons in Zugleitbetrieb!)

One thing I don't think we've mentioned yet is European Train Control System (ETCS) which presumably is ultimately intended to sort out all these variations across Europe.

In the UK we have ETCS on only one line (as far as I'm aware) - the Cambrian line, which is a rural line through Wales, with fairly low traffic density. It was deliberately chosen to trial the system without putting too much risk into the main network.

I'm no expert on ETCS but it's a cab signalling system rather than a fixed signal system. I understand it is also used in parts of Germany as well as other European countries.


James

Posts : 61
Join date : 2011-08-22

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by Northern Line on Sat Jun 23, 2012 10:49 am

When you say in cab signalling system is the CAWS on the same levels?
avatar
Northern Line

Posts : 281
Join date : 2011-07-12
Age : 23
Location : Tooting, London, England

http://www.bve-district.co.uk/

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by Quork on Sat Jun 23, 2012 12:19 pm

ETCS is no signalling system, it is a train protection system. There are several levels. I don't know them by heart as it's still far in the future; what I know, is, there are several current systems accepted as level 0, PZB and LZB both qualifzy for that. ATBL in Belgium and KVB in France probably as well, while british AWS or polish SHP probaby aren't. I'll check somewhere in the new week, now I'm at a friend's so don't want to spend too much time with research about railway topics.

The most simple ETCS system is Eurobalises at distant and at main signals, transmitting information about the signals aspect to the vehicle, just like PZB or ATBL. On high traffic density routes a Euroloop lies in front of the main signal (from about I think 200m before till the signal itself) and constantly transmits information about the signal aspect; this way an upgrad of the signal aspect results in an immediate upgrade of supervision speed in the train and thus an earlier acceleration as opposed to only balises, where the supervision is active till the main signal's balise is passed, upgrading the onboard information. The signal itself is a native signal of the country, so a three- or four-lamp light in the UK, a Ks or H/V or Hl signal in Germany or a christmas tree-style signal in Switzerland...

The highest levels of ETCS are in-cab signalling. This is independant from signals; there can still be signals to enable continuation of traffic in the event of ETCS failure, but there needn't be any. And as long as you're travelling on in-cab signalling, signals are of no interest to you (unlike irish CAWS). The in-cab signalling can be realized in different ways; there can be Eurobalises and Euroloops for information transmission, or the transmission is realized by GSM-R and Eurobalises are only there for positioning. Also here, if there are signals on the route, they're still the native signals of the given country, that's nothing to do with ETCS. Well, nearly; there are the so-called ETCS stop markers which need to be integrated into the signalling directives of the countries. In Germany, they're in directive 301 (signalling) as Ne14 since last December.

If you don't have plans for your summer vacation and want to see ETCS in action on a greater scale, go to e.g. Switzerland (several routes operated by infrastructure operators SBB and BLS), Italy (high speed routes Roma-Napoli and Milano-Turino), Letzebuerg (virtually the whole country is in ETCS level 1), Saudi Arabia (Riad-Dammam in ETCS level 1) etc.
avatar
Quork

Posts : 1087
Join date : 2012-05-05
Age : 26
Location : Hofheim a.T., Hessen (Hesse), European Union

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by James on Sat Jun 23, 2012 4:43 pm

Certainly in UK terminology ETCS (at its higher levels) is both a signalling system and a train protection system. There is no need for fixed signals (other than as a back up if you want it, as you say) so it is signalling the trains as well as providing train protection. I suppose the only problem with going to see it in operation is that unless you're in the cab there's not a lot to see! A trip to see some of that traditional signalling you provided pictures of sounds more tempting. Very Happy

Northern Line - as far as I understand its operation, CAWS is not on the same level as a high level ETCS implementation. All CAWS does, it seems, is indicate in the cab the aspect of the next signal, and require the Driver to acknowledge a warning tone (by pressing a button) if the aspect displayed is more restrictive than the previous one.

It's certainly not a signalling system (it is wholly reliant on the fixed signals) but even as a train protection system it doesn't supply full protection. If the Driver acknowledges the warning tone indicating the more restrictive aspect, he can still completely disregard it in terms of train control and go past a red signal. CAWS will not intervene.

The problem with systems like this (AWS is the same) is that pressing the 'acknowledge' button simply becomes an automatic response to hearing the warning tone. After a while it's very easy to do it entirely sub-consciously and therefore it doesn't do much to prompt the driver to react in controlling the train. It's not useless by any means, but it's a long way from being fool-proof.

It's because of this that TPWS was introduced in the UK. TPWS is pretty primitive and is irritating in its lack of flexibility, but it does overcome the problem of sub-conscious acknowledgement because it is not a system the Driver interacts with under normal circumstances.



James

Posts : 61
Join date : 2011-08-22

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by Quork on Sat Jun 23, 2012 5:47 pm

That's why the PZB sound comes after acknowledging. You see a signal you need to acknowledge, you press PZB Wachsam, the sound comes (as long as you press the button) and a braking curve is supervised. If you don't keep below, you're stopped.

In-cab signalling means, speaking of LZB or ETCS, you are shown the next speed and the distance till there and the current maximum speed. For LZB, that's
v(soll) - Maximum speed now
v(ziel) - target speed, e.g. 250 (in km/h of course) or 000 (for stop)
ZE - target distance, max. 13000 (metres) with LZB CIR ELKE (Computer Integrated Railroading zur Erhöhung der Leistungsfähigkeit im Kernnetz der Eisenbahn - yeah it really is that English/German mix...

Let's clarify the term "signalling system"; for me it is the equivalent to german term Signalsystem which means a set of defined aspects with defined meanings; e.g. "one or two reds" = Hp0, "one green" = Hp1, "green over yellow" = Hp2 (and corresponding distant signal aspects) for H/V. In this sense, ETCS isn't a signalling system. As an already real example; LZB is a train protection system and no signalling system, although it also makes signals obsolete.
avatar
Quork

Posts : 1087
Join date : 2012-05-05
Age : 26
Location : Hofheim a.T., Hessen (Hesse), European Union

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by James on Sun Jun 24, 2012 9:15 am

Having an acknowledgement without any sound to prompt it is an interesting system. I'd imagine the time window in which you must acknowledge must be longer than with AWS (which is 2.7 seconds, or 2 seconds for trains capable of exceeding 100mph). PZB sounds much more sophisticated than AWS or TPWS though, neither of which are anything like complex enough to supervise a braking curve.

I think in the UK the definition of 'signalling system' is a bit broader. It is basically just the system by which the trains are regulated and kept apart - it doesn't necessarily need to have defined aspects (red, green etc) as the output. Train protection is an optional extra to a signalling system, which in some way intervenes if the Driver doesn't react to the output of the signalling system.

I think the difference in our definitions is very small though.

James

Posts : 61
Join date : 2011-08-22

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by Quork on Sun Jun 24, 2012 11:17 am

It's 4 seconds.
avatar
Quork

Posts : 1087
Join date : 2012-05-05
Age : 26
Location : Hofheim a.T., Hessen (Hesse), European Union

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by Northern Line on Sun Jun 24, 2012 3:40 pm

Don't forget that TPWS also works alongside other restrictions as well as signalling, such as when you are warned of a restrictive speed which is lower than the normal line speed in this case let's say 40mph when the original line speed is like 60mph, you have to slow down to that restrictive speed usually after 15 seconds of passing it or whenever it is suitable for you to. If you were to acknowledge the AWS alarm and continue at that speed before coming to the restriction sign, you may be likely to pass over two TPWS grids without noticing and suddenly slow down to a stop. You also find the brake demand has been applied although you have not passed any SPADs or signals. This is because the warning further back up the line has been acknowledged but the speed had not been slowed to the restrictive. an example of this is on the approach to Salford Bridge station on the Salford loop in the route NWM.

Many new bve users don't tend to look at this sort of thing, most new users understand it but others don't. Even I never noticed when I first played the NWM I couldn't understand why I was stopped even though I was well ahead of the restrictive speed sign.
avatar
Northern Line

Posts : 281
Join date : 2011-07-12
Age : 23
Location : Tooting, London, England

http://www.bve-district.co.uk/

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by Quork on Sun Jun 24, 2012 6:39 pm

Yeah, but that's only punctual supervision. With PZB you have continuous supervision for 1250m after a 1000Hz beacon (at signals announcing a speed of 90km/h or less or a stop) and 250m after a 500Hz beacon (about 200m before a signal at danger or allowing 30km/h or less [former GDR: 40km/h or less] and in special situations), which is safer than testing only specific points.
avatar
Quork

Posts : 1087
Join date : 2012-05-05
Age : 26
Location : Hofheim a.T., Hessen (Hesse), European Union

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by James on Sun Jun 24, 2012 6:53 pm

TPWS was intended to be a system which did not require any alteration to driving style. In theory, if Drivers carried on driving as they always did, they would never notice or have to interact with the system unless they made a mistake.

In practice it didn't happen like that. It was soon found that Drivers were braking at their usual point, knowing that they would have speed down to the right speed by the commencement of the Permanent Speed Restriction (PSR) and were receiving a TPWS brake application. The problem was this: TPWS has two settings; freight and passenger. This is because passenger trains can reduce their speed more quickly under braking so with a passenger train you need to be able to pass over the TPWS grid at a higher speed than you could with a freight train.

The problem is that not all passenger trains have the same brake force available as each other, either. TPWS was set up on calculations based upon a train of traditional tread-braked stock. Modern disk-braked trains could brake harder and later, passing the TPWS grid more quickly but still able to be down to the correct speed for the commencement of the PSR. When Drivers of these trains did this, they got TPWS activations. Basically the TPWS system believed they were going to fast because it assumed they had less braking capability than they actually did.

This problem has never been solved. Drivers simply have to work around it by braking earlier and lighter than is otherwise necessary. This has some advantages - such as a smoother ride - but it reduces the ability of the Driver to make up time after delays. Nowadays most train companies have the speeds at which TPWS grids are set posted at traincrew depots. In other words, the TPWS grid has effectively become another permanent speed restriction for the Driver to comply with.

On BVE you get the same experience. It's a good idea to learn which Permanent Speed Restrictions are protected by TPWS grids, and you learn the hard way roughly what speeds they are set at! Personally I'd like to see route developers publish the speeds their TPWS grids are set at.





James

Posts : 61
Join date : 2011-08-22

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by Quork on Sun Jun 24, 2012 7:22 pm

There's a unit for measuring the braking ability of a train. It is called brake percentage. A vehicle has a brake percentage of 80, if an EB application at 120km/h leads to a stop after 1000m. (That's the definition). Each train has its brake percentage calculated before starting and after changes to the trainset. According to the brake percentage, the PZB computer chooses the correct programme and thus the correct brake curves. There are the programmes O (upper), M (middle) and U (lower). The threshold percentages are: 111 and more -> O, 66 and more -> M, 65 and less -> U. O has a max speed of 160km/h, M 120km/h, U 100km/h. The brake curves are steeper the higher the programme is. To take 1000Hz beacons as an example, it's (keep in mind the checking speeds are always 5km/h more than you're permitted, as the slightest touching of the brake curve leads to an immediate penalty braking)
165 -> 85 within 23sec for O
125 -> 70 within 29sec for M
100 -> 55 within 38sec for U

That's how the differences you mentioned are solved in PZB.
avatar
Quork

Posts : 1087
Join date : 2012-05-05
Age : 26
Location : Hofheim a.T., Hessen (Hesse), European Union

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by James on Sun Jun 24, 2012 8:35 pm

Like I say - PZB sounds a lot more sophisticated!

James

Posts : 61
Join date : 2011-08-22

Back to top Go down

Re: Traffic Lights and Multi-Aspect Signals

Post by Sponsored content


Sponsored content


Back to top Go down

View previous topic View next topic Back to top


 
Permissions in this forum:
You cannot reply to topics in this forum