• Our booking engine at tickets.railforums.co.uk (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!

ATO on Thameslink (?) and December 2019 timetable discussion

Status
Not open for further replies.

Bald Rick

Veteran Member
Joined
28 Sep 2010
Messages
29,070
How often do things actually go wrong in the core though - as in, a train coming to a halt and then everything getting stuck stacking up behind it?

I know everything's sensitive to disruption elsewhere, which can cause trains to enter the core out of sequence and therefore not as efficiently as would be ideal - however I was under the impression that something breaking down in the core is now relatively rare.

More often that it should, but less often than many people think. It’s rare to have anything stopped for more than 10 minutes.



If that is so then the passenger flows on and off the train are much faster than expected, which is good, of course and does suggest that more than 24ph as a timetabled service might be possible. But can I ask some questions (don't get offended, I really want to know):

1) Is that wheels stop to wheels start?
2) Loading conditions (45s in recovery conditions implies a crush load)
3) Which station(s)?
4) AM or PM peak?
5) Mainly boarders or alighters?

1) yes
2) variable
3) all of the core stations
4) both, and off peak
5) both

You can only see this at times when the train is late, as they are normally booked 1 minute dwell. ‘Recovery conditons’ doesn’t imply crush loading though, it can be as simple as one to two trains out of sequence, requiring trains to be pumped through at 120-150sec headways to catch up. For example two trains arriving 10 minutes late right behind each other will take 8 successive 120sec headways to recover at 24 tph.

I have often seen trains through on 120second headways, but this requires sharp dispatch and sharp driving; the latter of which will be universal when ATO is in use.


The trains were specifically designed to deal with several hundred boarders and alighters in a 45sec wheels stop / Wheels start, experiments were conducted nearly a decade ago to prove it.
 
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R

RailUK Forums

ComUtoR

Established Member
Joined
13 Dec 2013
Messages
9,398
Location
UK
1) Is that wheels stop to wheels start?
2) Loading conditions (45s in recovery conditions implies a crush load)
3) Which station(s)?
4) AM or PM peak?
5) Mainly boarders or alighters?

Stop train at stop mark
About 3 seconds before I pop the door button(s) (ato should cure that delay ?)
All doors open pretty fast (If I hit the mark correctly - I've only missed once :oops:)
Check schedule (depending on departure time or booked dwell time I aim to close 30 seconds before departure. This is achieved at Pancras southbound as you can arrive early on some services and there is timed recovery at Pancras and City you can gain half a minute easy)
Loading and unloading (easily within 30 seconds on a normal day but you can do it around 15 seconds. When its busy your pushing 30-45 and then I start getting a bit anxious. When it goes up the creek you can push a minute and a bit. 700s swallow passengers)
PTI check (give it a little over a second per monitor and then hit the button(s). The doors will close within 10 seconds and when there is an opportunity that some may have already closed (auto close function) they click almost instantly)
Train safety check (Required final check again taking around a second per monitor)
Check signal (about a seconds glance)
Take power

The 24tph does have recovery built in (I think) but the current timings will be tightened up a little.
 

DerekC

Established Member
Joined
26 Oct 2015
Messages
2,107
Location
Hampshire (nearly a Hog)
Thanks @Bald Rick and @ComUtoR for the information which is very helpful. Does either of you know if GTR or NR is analysing the dwell time performance to see how the modelling and analysis at design stage (which I am well aware of) is working out in practice? I have a genuine industry-beneficial reason for asking which I will explain via a PM on request.
 

ComUtoR

Established Member
Joined
13 Dec 2013
Messages
9,398
Location
UK
I just drive them :) I leave the technicalities to our resident expert Mr B. Rick.

Each TOC does or should have a performance department. Mine is currently working towards RTR (Right time railway) and is analyzing every aspect of the entire process. Dispatch is a little bit of a fight between Drivers, Dispatch Staff and Performance. The tighter you make it, the more you increase the chance of a delay. There needs to be a minimum allowance and then let the people on the front line compensate.

I'd rather have a robust pathway so that I can arrive on time. If I can make it a little earlier then I can allow a more relaxed dispatch procedure where dwell times may be tight. If my pathway is tight but my dwell time is long it often defeats the object. If I run in late it always cuts into my dwell times so there is no real benefit.

The major issue is of course....... PEOPLE. I get to people watch all day. I have seen the dumbest of human behaviors. I don't think you can model for that. The addition of the loading areas and the loading screen on the platform screens I'm not sure about. I'd like to think they help but where people load and alight from is deeply rooted in where the passenger wants, not what is best for loading. You could print out the HMI screen door pattern and I'm sure Drivers could tell which station is is based on which doors open.

As the PTI gets more and more focus the rules are being tightened up and staff are more and more risk adverse. Any modelling needs to take that into account. Will ATO allow for the doors to be reopened or when the unit gets interlock will it just pull away ? The rules say there must be a final train safety check after interlock is gained. That will still need to be done by the fleshy part of the system. If it doesn't, I can guarantee that Drivers will not close the doors until they are 110% confident that there is nobody attempting to board, boarding, running down the platform, crowding the doors, etc, and that will almost eliminate any gains.
 

Need2

Member
Joined
15 Jun 2016
Messages
595
Although it probably will not happen, it would be interesting to see what happens with train dwell times and timings if the recent RAIB recommendation of a 13 second 'final safety check' AFTER door interlock is gained is implemented!
(recommended after the recent trap and drag of a dog at Elstree)
 

DerekC

Established Member
Joined
26 Oct 2015
Messages
2,107
Location
Hampshire (nearly a Hog)
Will ATO allow for the doors to be reopened or when the unit gets interlock will it just pull away ? The rules say there must be a final train safety check after interlock is gained. That will still need to be done by the fleshy part of the system. If it doesn't, I can guarantee that Drivers will not close the doors until they are 110% confident that there is nobody attempting to board, boarding, running down the platform, crowding the doors, etc, and that will almost eliminate any gains.

From memory common sense was applied to the design and there is a "depart" button (or similar) which you press after door interlock has been gained and the final safety check done by the fleshy part of the system! You can also recycle the doors if needed before pressing "depart". I think that if you are in ATO the cameras stay on as the train starts to roll, but I am not sure about that bit.
 

DerekC

Established Member
Joined
26 Oct 2015
Messages
2,107
Location
Hampshire (nearly a Hog)
Although it probably will not happen, it would be interesting to see what happens with train dwell times and timings if the recent RAIB recommendation of a 13 second 'final safety check' AFTER door interlock is gained is implemented!
(recommended after the recent trap and drag of a dog at Elstree)

Yes, I wondered about that. A fixed time doesn't seem sensible to me. It takes as long as it takes for the driver to be happy it's safe to depart.
 

ComUtoR

Established Member
Joined
13 Dec 2013
Messages
9,398
Location
UK
From memory common sense was applied to the design and there is a "depart" button (or similar) which you press after door interlock has been gained and the final safety check done by the fleshy part of the system!

That isn't a button I have seen and certainly isn't a physical button in the cab. The HMI (in a 700) has plenty of spaces for software buttons but I can't confirm/deny any 'depart' button. That would be an interesting and important addition.

You can also recycle the doors if needed before pressing "depart".

As long as there is some form of delay before the unit pulls away or there is some kind of confirmation that dispatch is complete. Recycling doors needs to be prevented as any restart of the process again, eliminates any gains made through ATO.
 

ComUtoR

Established Member
Joined
13 Dec 2013
Messages
9,398
Location
UK
Yes, I wondered about that. A fixed time doesn't seem sensible to me. It takes as long as it takes for the driver to be happy it's safe to depart.

It is a minimum allowance. Not forgetting that they have 'discovered' that Drivers were pulling away within 5 seconds. If it was our decision then that would have been considered to be acceptable. However, the RSSB saw fit to state that this isn't sufficient.

As per my example. I have to allow 3 seconds for operational checks once I arrive before popping the doors open. I am required to allow 'sufficient time for boarding (approx 15 seconds), carry out the first train safety check (potentially 13 seconds) gain interlock, and then the second train safety check (another 13 seconds)

We can do it faster and there is overlap in the way humans think and act but that isn't being allowed for because of the clinical and mechanical approach the RSSB takes with everything.
 

ComUtoR

Established Member
Joined
13 Dec 2013
Messages
9,398
Location
UK
ComUtoR as a point of clarity do you have any interaction with platform staff in the core for despatch?

Not anymore. It's DOO. Professionalism and experience goes a long way as they may flag you down for a wheelchair or you have a gander as your stopping checking for any dispatch requirements. Other than that, generally they don't exist.
 

Bald Rick

Veteran Member
Joined
28 Sep 2010
Messages
29,070
Will ATO allow for the doors to be reopened or when the unit gets interlock will it just pull away ?

That isn't a button I have seen and certainly isn't a physical button in the cab. The HMI (in a 700) has plenty of spaces for software buttons but I can't confirm/deny any 'depart' button. That would be an interesting and important addition.

There’s an ‘ATO START’ button, or at least there was on the mock ups and simulator. Once you have interlock and done the check, press that and off you go. It’s a physicals button, right hand side of desk.

Thanks @Bald Rick and @ComUtoR for the information which is very helpful. Does either of you know if GTR or NR is analysing the dwell time performance to see how the modelling and analysis at design stage (which I am well aware of) is working out in practice? I have a genuine industry-beneficial reason for asking which I will explain via a PM on request.

I’m afraid I don’t know of anything recent, but then I’m not involved.
 

ComUtoR

Established Member
Joined
13 Dec 2013
Messages
9,398
Location
UK
There’s an ‘ATO START’ button, or at least there was on the mock ups and simulator. Once you have interlock and done the check, press that and off you go. It’s a physicals button, right hand side of desk.

Coolio. Fears assuaged then. There was an assumption (yes they are bad) that ATO start was the start process then you sit back and wake up at Kentish, then press ATO stop. If it's 'start' each time then I'm a lot happier. (no training; as you can tell)

Cheers BR.

Now I'm off to the football. Keep up the debate, tis getting interesting :)
 

carriageline

Established Member
Joined
11 Jan 2012
Messages
1,897
It is Bi-Di under traditional signals. Over ATO, I don't know.

My concerns are how quickly trains stack up and how easy the system snaps. You can literally be on time 1 signal outside of Blackfriars and then drop 5/6 minutes on a normal day and when it really goes up the creek you can drop loads. When your sitting on the last signal out of Blackfriars your options are limited. When you have trains over junctions and crossovers it just compounds the issues further. I remember an incident this year where there was a dragging brake at Blackfriars and everyone stacked up. I could have easily stayed at Elephant and passengers emptied out onto the Tube. Once I passed the junction signal it was game over and my only option is to queue up for The Core :/

Some days I despair with The Core. I would rather be held in a station with the crossover in front held clear but it doesn't happen. You start stacking up really quickly. I'd like to know how ATO will deal with disruption. How easy will it be to turn back at Kentish Town or drop into Smithfield and turn back there ? Will the system revert to legacy and then the human in the box figure it all out or will it be fed into the ATO/ERTMS and the system come up with a recovery plan ?

It's stuff like this where I'm interested in how ATO will benefit/cure/help/ameliorate the situation. If everything is forced into time order your potentially waiting for something behind you to be regulated in front.

I know Bald Rick summed it up, but I’ll provide some techy detail. ERTMS is essentially just a signalling system (in the context we are discussing at least). ATO is the software on the train that allows it to drive itself.

ETCS will tell the train how far it can go, and at what speed and where it must stop. The train (using that information) works out its braking and acceleration curves. Using all that information, ATO will drive the train. So ATO doesn’t really make any decisions it self, it just stops, starts and opens doors. The signaller sets a route (identical to today) and the train is told how fast and where to stop.

Your Smithfield scenario will require the signaller setting the route, the train goes into “On Sight” mode, and driver manually drives it in (as it’s permissive). If you want more info on all this ETCS/ATO stuff, drop me a message.

The driver then drives the train back out conventionally as ETCS is then not available for trains leaving Smithfield’s.

Under ETCS/ATO nothing really changes. The signallers and control make the decisions like they do today. The only difference is ATO will drive the train as it’s told.

What your discussing is Traffic Management, which will feed into all the ATO/ARS etc. If you want more info on that, drop me a PM. I can’t really go into too much detail public ally.

On the topic of Core fragility.. The problem now is Thameslink covers such a huge area. An issue in Kent will start to smash trains heading into the core. These trains then start hitting trains on the ECML/MML, and delays snowball very fast. I think it’s made the core a lot more fragile, as your reliant on much more infrastructure behaving.
 

DerekC

Established Member
Joined
26 Oct 2015
Messages
2,107
Location
Hampshire (nearly a Hog)
That isn't a button I have seen and certainly isn't a physical button in the cab. The HMI (in a 700) has plenty of spaces for software buttons but I can't confirm/deny any 'depart' button. That would be an interesting and important addition. As long as there is some form of delay before the unit pulls away or there is some kind of confirmation that dispatch is complete. Recycling doors needs to be prevented as any restart of the process again, eliminates any gains made through ATO.

There’s an ‘ATO START’ button, or at least there was on the mock ups and simulator. Once you have interlock and done the check, press that and off you go. It’s a physicals button, right hand side of desk. I’m afraid I don’t know of anything recent, but then I’m not involved.

That makes sense - my memory was wrong about what it is called.

Coolio. Fears assuaged then. There was an assumption (yes they are bad) that ATO start was the start process then you sit back and wake up at Kentish, then press ATO stop. If it's 'start' each time then I'm a lot happier. (no training; as you can tell)

HEALTH WARNING - THE FOLLOWING MAY BE OUT OF DATE SO PLEASE TREAT WITH CAUTION:

To engage ATO (assuming you are in ETCS Level 2) you put the PBC to neutral and press the "ATO START" button. You can do that either stationary or on the move. Once in ATO the train will keep moving under automatic control until it stops at a station - then you press the "ATO START" button again once the doors are closed. If the train is in ATO and stops at a red signal/end of movement authority it will start to move again automatically once the movement authority is extended. There should also be an "ATO STOP" button which causes a full service brake application - once the train has stopped you would have to press "ATO START" again to move off. You can also take over simply by moving the PBC out of the neutral position - you are immediately back in manual control.
 

Aictos

Established Member
Joined
28 Apr 2009
Messages
10,403
That makes sense - my memory was wrong about what it is called.



HEALTH WARNING - THE FOLLOWING MAY BE OUT OF DATE SO PLEASE TREAT WITH CAUTION:

To engage ATO (assuming you are in ETCS Level 2) you put the PBC to neutral and press the "ATO START" button. You can do that either stationary or on the move. Once in ATO the train will keep moving under automatic control until it stops at a station - then you press the "ATO START" button again once the doors are closed. If the train is in ATO and stops at a red signal/end of movement authority it will start to move again automatically once the movement authority is extended. There should also be an "ATO STOP" button which causes a full service brake application - once the train has stopped you would have to press "ATO START" again to move off. You can also take over simply by moving the PBC out of the neutral position - you are immediately back in manual control.

When the train in question is in ATO mode, do you still have to press the DSD pedal with your foot like you do with non ATO trains to stop the brakes from applying automatically?

Or is it a case of once the train is in ATO mode then the driver doesn't do anything other then press the ATO Start button to move from stations once the doors have closed?

I trust also once in ATO mode, the train arrives at Farringdon then stops and releases the doors automatically with no input from the driver after a few seconds closes them again no input from driver then the driver presses ATO start and off you go or am I'm misunderstanding ATO in the Core?
 

Meerkat

Established Member
Joined
14 Jul 2018
Messages
7,419
Interested in drivers’ opinions as the 1 second per monitor sounds silly - a second is actually a long time in vision terms (instinctively I reckon I would check all three car mirrors and over shoulder in that time, and always surprised how much football there is before a ‘scored in ten seconds’ goal)
If you spend a second on each monitor that is a long time for something to happened on the first monitor whilst you check the rest, or is that you repeatedly check all monitors but overall it should take 1 second multiplied by the number of monitors?
 

ComUtoR

Established Member
Joined
13 Dec 2013
Messages
9,398
Location
UK
Carriagline and Derek - My thanks.

Meerkat...

A second isn't enough and is enough if you catch my drift. We are talking about hundreds of people going through 16/24 sets of doors in around 15-30 seconds. You have people crowding on the platform, people, bags, animals all hanging out in the doorway, clothing and fingers you just can't see and people who still refuse to accept that the 'Hustle alarm' means the doors are CLOSING and they should step back, not suddenly dive through it.

That said, a single second per monitor in a bank of 8 leads to errors. The problem is that Drivers are skimming through and only glancing each monitor and not accounting for the details. They are missing a coat getting caught or a dog lead or a handbag. You are right in the fact that by the time you have checked all 8 then something else happens in the first one :/ The RSSB advice stated (IIRC) that it should take 13 seconds for an 8 car and whilst that is reasonable it still fails to allow for operational factors. Which should allow for platform based monitors and onboard cameras.

You are also right that 1 second can be a long time but we are staring into monitors all day, every day. The impact that has on a Drivers workload is considerable so they need to advice enough time is given to ensure a dedicated and robust check is made. I think the report stated that Drivers were pulling away at 3.3 seconds or something silly like that. That just isn't enough for a true safety check. A glance is not proving to be safe. It is the Gorilla scenario :)

There is also the issue of the monitor quality and camera placement. The DOO monitors in a 700 are tiny, considering what they are for. The still wash out in the sun, they go pink/blue/green or the just keep blinking out. That also needs an allowing for. You are also still fighting against infrastructure issues. Granted you only need the small space in the doorway but somebody can be standing right next to a door and still be invisible on the camera and because it only sees people really close to the train, any tight platforms with curvature or walls. Peckham Rye is a £!£"$££$ nightmare for that.

I think the advice they have given is a game changer. Not forgetting that this needs to be applied twice per dispatch. Train safety check before closing, and the final check after interlock is gained. It will be interesting to see if they only apply it for the check after interlock.
 

Meerkat

Established Member
Joined
14 Jul 2018
Messages
7,419
Thanks for the response. Doing it so thoroughly twice seems overkill.
If the second one is a thorough check (which should be easier as most people will step away from a closed door?) then surely the first is fairly cursory - what are you checking for, someone who won’t be able to move as the doors close? Passengers should take the responsibility to be clear of the closing doors if the driver is then checking on them once they are closed.
 

ComUtoR

Established Member
Joined
13 Dec 2013
Messages
9,398
Location
UK
Thanks for the response. Doing it so thoroughly twice seems overkill.

You have to carry out a safety check BEFORE closing the doors and the rules got tightened up recently. If you just glance it and trap someone in the doors then that is the first step to a more serious incident. Personally I think the more important check comes first. You need to ensure the doors are clear and nobody will get trapped.

If the second one is a thorough check

I think the RSSB response is reactionary to the incident(s) If a through check took place after interlock then the Driver wouldn't pull away and the doors could be re released. The evidence they have gathered lends itself to having a final check as that is where the incident occurs and could be prevented. Not forgetting that due to previous incident, interlock is no longer acceptable as a final check the doors are clear. But because it is a reaction to incidents, they haven't considered the entire process.

(which should be easier as most people will step away from a closed door?)

Errr... Um...... I wish they would.

then surely the first is fairly cursory - what are you checking for, someone who won’t be able to move as the doors close?

The first check has more of a decision process involved. Its not just about if the doors are clear but will that drunk girl on the platform approach the train as it moves off or will that old person suddenly remember that this is her train and stick her cane in the door or will that commuter run up the stairs or down the ramp and dive in at the last second. Our PDP (professional driving policy) states quite clearly that you need to ensure that you have allowed for vulnerable people who may be attempting to board or alight before you close the doors. Places like Peckham are so tight that you have no clue who is boarding or alighting. There is a lot of judgement involved. Hence why a cursory glance becomes dangerous and the first check is just as important (if not more) than the second.

The second check is purely to see if you done it right the first time and that nothing or nobody is trapped (because interlock cannot be trusted anymore) You will have already checked the platform and some doors and the platform surrounding them are potentially clear and you are only checking the specific doors where people were/are.

Passengers should take the responsibility to be clear of the closing doors if the driver is then checking on them once they are closed.

From a personal standpoint I'd agree. From a legal and operational one. I can't. The advice being given is to ensure that ample time is given at the last point in which an incident could occur.
 

Meerkat

Established Member
Joined
14 Jul 2018
Messages
7,419
I understand the driver view - if you are responsible then it takes as long as it takes, otherwise you want it in the rules that the passenger is responsible.
IMO the first check should be cursory - the hustle alarm is the mitigation. They are going for safER than is reasonably practicable! The underground just slams the doors on people and the percentage of incidents must be minuscule or there would be hundreds dying!
Hustle alarm, doors close. With a public information programme to highlight this. Anything else just encourages delay - “driver won’t close the doors if we get near enough...”
 

carriageline

Established Member
Joined
11 Jan 2012
Messages
1,897
Recently I watched someone jump out of their seat (inside the train), and strolled towards the door as they realised they was at the doors

Problem was, as the person jumped up the driver had already closed the doors. By the time the person reached the doors, the train was starting the move. The person started to jab the “open” button as if the doors would open.

At times, people’s common sense disappears, and with it their sense of “self safety”. It seems to be worse when people are running late, desperate to get places. Go stand on a tube platform during the peak, and watch the desperation of people trying to get on an already packed train!

If people behaved correctly 100% of the time the driver wouldn’t even need to do safety checks!
 
Status
Not open for further replies.

Top