• Our new ticketing site is now live! Using either this or the original site (both powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!

European Train Control System - Software issue End of Authority/Limit of Authority trip

bahnause

Member
Joined
30 Dec 2016
Messages
667
Location
bülach (switzerland)
I also agree with the above, unbelievably bad UI to have a green flashing button that you're not meant to press!
It is meant to be pressed without contacting the signaller according to the design of ETCS. It might be an odd implementation of ETCS on a certain line to make it desirable not to press it or they don’t trust the staff to identify the issue and act accordingly. The explanation given does make no sense to me.
 
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R

RailUK Forums

crablab

Member
Joined
8 Feb 2020
Messages
1,013
Location
UK
What's the purpose of having the driver read the error message to the signaller? In the example scenarios, the message shown was to the effect that the train had exceeded the EOA/LOA. That seemed to be implicit given the signaller had watched the train enter the next block without authority.

Perhaps other issues can cause a trip? I presume things like the train losing positional information or connection to the RBC, as examples? In those scenarios, maybe the flow chart is more use?
 

Fincra5

Established Member
Joined
6 Jun 2009
Messages
2,584
What's the purpose of having the driver read the error message to the signaller? In the example scenarios, the message shown was to the effect that the train had exceeded the EOA/LOA. That seemed to be implicit given the signaller had watched the train enter the next block without authority.

Perhaps other issues can cause a trip? I presume things like the train losing positional information or connection to the RBC, as examples? In those scenarios, maybe the flow chart is more use?
A passing of an EoA or LoA is one of couple of reasons where a TRIP could occur from driver error. The majority of TRIPs are system related.

Thus the importance of the TRIP message so it can be identified if the TRIP was driver error (akin to a SPAD) or system side.
 

bengley

Established Member
Joined
18 May 2008
Messages
1,931
What a nonsense system. Give me lights on sticks any day over all this rubbish!
 

GN Boy

Member
Joined
8 Dec 2020
Messages
92
Location
England
What a nonsense system. Give me lights on sticks any day over all this rubbish!

Ironically, the “lights on sticks” were initially the go-to get out of trouble for NCL journeys that encounted problems with ETCS!
 

Jonny

Established Member
Joined
10 Feb 2011
Messages
2,573
Surely, just in general terms of the information system, the signaller should know or be able to find out whether or not the train has gone beyond/outside a movement authority (especially if it is SPAD-equivalent) without the need for any information from the driver?
 

choochoochoo

Established Member
Joined
6 Aug 2013
Messages
1,252
For those unaware of it, the user interfaces to ETCS are bespoke and not standardised.

However, I’m not aware of how many supplier options there are for the drivers’ interface kit.

Does that mean they could’ve chosen a different colour arrow/indicator to mark the EoA on the planning screen on the DMI ? Why make it the same colour as speed changes ? A bright red EoA arrow would be more conspicuous than the dull grey one that currently sits at the EoA.

What's the purpose of having the driver read the error message to the signaller? In the example scenarios, the message shown was to the effect that the train had exceeded the EOA/LOA. That seemed to be implicit given the signaller had watched the train enter the next block without authority.

Perhaps other issues can cause a trip? I presume things like the train losing positional information or connection to the RBC, as examples? In those scenarios, maybe the flow chart is more use?

These trains communicate directly to the signalling system. Why isn’t the trip message being sent directly to the signaller when a trip occurs ? Why does a driver need to tel, the signaller what is on a screen ?
 

TurboMan

Member
Joined
5 Apr 2022
Messages
411
Location
UK
Does that mean they could’ve chosen a different colour arrow/indicator to mark the EoA on the planning screen on the DMI ? Why make it the same colour as speed changes ? A bright red EoA arrow would be more conspicuous than the dull grey one that currently sits at the EoA.
No - the DMI layout, icons, colours etc. are defined by the standard (ERA_ERTMS_015560 if you're interested). However, the EoA 'flag' on the planning area is not the only indication of the EoA - there is a 'distance to target' indicator on the left, and the train will be in target speed monitoring, with the 'hook' on the circular speed gauge dropping towards zero as the train approaches the EoA.

Perhaps other issues can cause a trip? I presume things like the train losing positional information or connection to the RBC, as examples? In those scenarios, maybe the flow chart is more use?
Yes, lots of reason for the train entering trip mode (TR) - about two dozen in total according to the standard, including such things as the train receiving a 'stop in SR' command from a balise group when in Staff Responsible (SR) and the override isn't active, or a 'stop in SH' command if trying to leave a depot while still in Shunt mode (SH).

However, loss of RBC connection isn't one - there is a brake application that occurs a certain time after the connection is lost (the time varies according to where you are on the network!), but this isn't the train actually entering TR.
 
Last edited:

choochoochoo

Established Member
Joined
6 Aug 2013
Messages
1,252
No - the DMI layout, icons, colours etc. are defined by the standard (ERA_ERTMS_015560 if you're interested). However, the EoA 'flag' on the planning area is not the only indication of the EoA - there is a 'distance to target' indicator on the left, and the train will be in target speed monitoring, with the 'hook' on the circular speed gauge dropping towards zero as the train approaches the EoA.
Whoever came up with those standards clearly didn’t talk to many drivers.

Distance to Target and the 'hook' aren’t as good for situational awareness as the planning area. Would it be so hard to have specified different colours for different targets ?
 

choochoochoo

Established Member
Joined
6 Aug 2013
Messages
1,252
They did, but mostly to German drivers. The distance to target and hook are almost identical to their existing LZB interface.
Looks like human factors getting ignored on the railway once more. If it’s not cab ergonomics it’s software design.
 

D365

Veteran Member
Joined
29 Jun 2012
Messages
12,148
Looks like human factors getting ignored on the railway once more. If it’s not cab ergonomics it’s software design.
HF is heavily involved in the cab redesigns but, aside from some hardware modifications to the driver displays, they don’t appear to have any influence over the software interface.

Plus - even ignoring all of the prescribed user interface requirements and regulation - a lot of ETCS software is passed to us from Germany with little-to-no means of making any adjustments.
 

choochoochoo

Established Member
Joined
6 Aug 2013
Messages
1,252
HF is heavily involved in the cab redesigns but, aside from some hardware modifications to the driver displays, they don’t appear to have any influence over the software interface.

Plus - even ignoring all of the prescribed user interface requirements and regulation - a lot of ETCS software is passed to us from Germany with little-to-no means of making any adjustments.
The cabs of a 700 or 717 suggest HF wasn’t considered. Both are hideously thought out cab layouts. Either that or the DfT spec was terrible and Siemens got off easy with what they provided on the cheap.

It does look like it’s a case of here’s the tools, and even though not ideal for the job here in the UK, you’ll make it work. It’s not ideal for transitioning from lights on sticks to in cab signalling. And even worse when you’re having to operate on both systems over different parts of the route.
 

Fincra5

Established Member
Joined
6 Jun 2009
Messages
2,584
The cabs of a 700 or 717 suggest HF wasn’t considered. Both are hideously thought out cab layouts. Either that or the DfT spec was terrible and Siemens got off easy with what they provided on the cheap.

It does look like it’s a case of here’s the tools, and even though not ideal for the job here in the UK, you’ll make it work. It’s not ideal for transitioning from lights on sticks to in cab signalling. And even worse when you’re having to operate on both systems over different parts of the route.
Eh? The 700 has one of the best laid out cabs I've driven from! Everything is in reach, set our clearly and with a good centralised driving position.
ETCS is a new way of driving but once you're used to it, you'll appreciate the benefits of a Class A safety system.
 

GN Boy

Member
Joined
8 Dec 2020
Messages
92
Location
England
Eh? The 700 has one of the best laid out cabs I've driven from! Everything is in reach, set our clearly and with a good centralised driving position.
ETCS is a new way of driving but once you're used to it, you'll appreciate the benefits of a Class A safety system.

I’d love for the 700 to have a holder/clip for your schedule card, though.
 

bahnause

Member
Joined
30 Dec 2016
Messages
667
Location
bülach (switzerland)
ETCS is a new way of driving but once you're used to it, you'll appreciate the benefits of a Class A safety system.
Unfortunately, it is not a class A safety system. We have four different ETCS suppliers with six different versions of onboard equipment, two different baselines and three software versions. It has become obvious, what an absolutely incoherent system has been created. Four lines with three different software versions complete the fun. In level 1LS there are braking curves and monitoring speeds that deviate so far from the actual achievable values of the rolling stock, that a lot of capacity on the network is lost. I‘m convinced once the pressure is high enough more and more of the european standards will be ignored and more “local versions“ will be created.
 

Fincra5

Established Member
Joined
6 Jun 2009
Messages
2,584
Unfortunately, it is not a class A safety system. We have four different ETCS suppliers with six different versions of onboard equipment, two different baselines and three software versions. It has become obvious, what an absolutely incoherent system has been created. Four lines with three different software versions complete the fun. In level 1LS there are braking curves and monitoring speeds that deviate so far from the actual achievable values of the rolling stock, that a lot of capacity on the network is lost. I‘m convinced once the pressure is high enough more and more of the european standards will be ignored and more “local versions“ will be created.

I'd say Automatic Train Protection at every signal/block is quite the increase in the UK compared to TPWS/AWS.

From my experience with two different implementations of the system, one which used ATO, it really pushes the limit of the train within Level 2 (we don't use Level 1).

Personally I prefer it these days compared to conventional signalling, you get to see your Speed Profile, Movement Authority for lots of KMs and it's designed to stop and incident occuring before it does.
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
18,559
As far as I know, there is no operational ETCS Level 1LS in the UK.

My understanding is that it is also a far more broad set of standards than the rest of the ETCS specification, and functionality can be produced that closely replicates an existing train protection system if desired.
See the Swiss installations that produced a functionally drop in replacement for Integra-SIGNUM.

EDIT:
Obviously not a railway engineer etc etc etc, but a naive reading of the messages available in a ETCS L1LS balise would indicate that a drop in replacement for both AWS and TPWS could be designed using standard ETCS hardware.
 
Last edited:

mrmartin

Member
Joined
17 Dec 2012
Messages
1,151
To be honest I find it astonishing that ECTS doesn't relay those error messages back to the control centre automatically as soon as they happen (alongside loads of diagnostic info which may be useful). This way the signaller would know what is up with the train before the driver even contacts them. Bizarre.
 

bahnause

Member
Joined
30 Dec 2016
Messages
667
Location
bülach (switzerland)
My understanding is that it is also a far more broad set of standards than the rest of the ETCS specification, and functionality can be produced that closely replicates an existing train protection system if desired.
See the Swiss installations that produced a functionally drop in replacement for Integra-SIGNUM.
As soon as braking curves in Level 1LS are involved - and they are in a lot of countries including switzerland - you are pretty much stuck with a lot of the original ETCS specs (braking cureves, release speeds and visualisation on the DMI). It is not a 1:1 copy anymore and the downfalls of the system become obvious. It is a really bad interface for the driver and safety standards at some points have to be lowered compared to the legacy system to keep capacity and timekeeping at previous levels. It is no longer the unobtrusive system that monitors the train driver in the background. It requires significantly more attention, especially in the delicate phases of the journey when attention should be focussed on the route and the signals.
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
18,559
As soon as braking curves in Level 1LS are involved - and they are in a lot of countries including switzerland - you are pretty much stuck with a lot of the original ETCS specs (braking cureves, release speeds and visualisation on the DMI). It is not a 1:1 copy anymore and the downfalls of the system become obvious. It is a really bad interface for the driver and safety standards at some points have to be lowered compared to the legacy system to keep capacity and timekeeping at previous levels. It is no longer the unobtrusive system that monitors the train driver in the background. It requires significantly more attention, especially in the delicate phases of the journey when attention should be focussed on the route and the signals.
My (admittedly limited) understanding is that the SIGNUM-Integra replacement installed in Switzerland (as opposed to the ZUB replacement, which has a Euroloop) uses a balise to impose an instantaneous speed limit on the train passing over the loop based on the signal state. If the train is going faster than the set speed, either 0 for stop or some speed for caution, then the train will trip the emergency brake.

In a TPWS or AWS replacement, couldn't you just set the speed to a precalculated value, analogously to the calculation used to position a TPWS "trigger" loop? That could be done with whatever braking curve you want couldn't it?
 
Last edited:

bahnause

Member
Joined
30 Dec 2016
Messages
667
Location
bülach (switzerland)
My (admittedly limited) understanding is that the SIGNUM-Integra replacement installed in Switzerland (as opposed to the ZUB replacement, which has a Euroloop) uses a balise to impose an instantaneous speed limit on the train passing over the loop based on the signal state. If the train is going faster than the set speed, either 0 for stop or some speed for caution, then the train will trip the emergency brake.
Signum is a very simple system based on magnets. It can therefore only distinguish three things due to the possible output (none, S-N or N-S): Free run (no input), warning or stop. It cannot monitor speeds. Only with tinkering solutions such as time-controlled switching of a magnet.
ZUB is also based on balises and optional loops, but is much more complex. Most of the SBB Network is/was equipped with ZUB additionally to SIGNUM. It sends data telegrams with target distance, target speed, gradient, line speed and other small details. This enables a continuous monitoring. The onboard device calculates a braking curve based on the train data (maximum speed, braking capacity, train length, braking type). An update can be sent out using an infill balise or a loop. The vehicle can only ever read the loop that was announced by a balise. Otherwise, loops in the opposite direction or from neighboring tracks would also be read. Loops are good for increasing capacity, but cannot influence a starting train. For this reason, balises are generally used at stations with starting or reversing trains, as they can stop an erroneously departing train before the danger point. However, this reduces capacity, as the upgrade of a signal aspect can only be sent to a train at certain points.
For rolling stock with Baseline 2, the SIGNUM magnets and ZUB balises have been replaced by ETCS balises which, however, transmit national data (ZUB data) in data packet 44. This is a 1:1 replacement, which replaces parts of the hardware. However, the ECTS Onboard Unit has nothing to do with these data, which are processed by the legacy hardware on the trains. No difference for the drivers (except for the technical stuff).
With Baseline 3, however, the ETCS telegrams are used and packet 44 is ignored. The ETCS standards also come into play here, i.e. braking curves and release speeds. ZUB does not have a release speed; the train driver must « free » himself if necessary and if permitted and not inhibited by the last telegram (<40 km/h). If he does this by mistake, hthe train can be stopped again by a balise or a loop. With ETCS, the release speed specified by the system is always automatically active and the train can only be stopped at the signal showing danger. With this system, a clearance behind the signal is therefore necessary or a very low release speed must be used. ZUB does not have this restriction. And this is also one of the main problems. You never know exactly when the monitoring of the braking curve will be replaced by the release speed and which release speed will be used (or even several in succession). But they have one thing in common: They are always active far too early. They require a driving style that deviates far too much from the performance the train is actually capable of. You look far too often at the display to see if a release speed is suddenly announced instead of concentrating on stopping the train in the correct place at the platform and before the signal.
In a TPWS or AWS replacement, couldn't you just set the speed to a precalculated value, analogously to the calculation used to position a TPWS "trigger" loop? That could be done with whatever braking curve you want couldn't it?
If I understand you correctly, this would be a step backwards from continuous monitoring to a punctual intervention. In addition, it would probably not be possible to monitor all permissible speeds according to train type and braking force. Depending on the rolling stock, these can be between 40km/h and 160km/h at the same point.
 
Last edited:

cool110

Member
Joined
12 Dec 2014
Messages
657
Location
Preston
If I understand you correctly, this would be a step backwards from continuous monitoring to a punctual intervention. In addition, it would probably not be possible to monitor all permissible speeds according to train type and braking force. Depending on the rolling stock, these can be between 40km/h and 160km/h at the same point.
It would be 1:1 with TPWS rather than a step backward, and a slight improvement in 4 aspect areas since unlike AWS it could distinguish between single and double yellow.
 

Top