DOO: what method do you prefer?

[Bletchleyite]

DOO early 1990s style….

And Swiss/German DOO - they rely on the interlock with just a wing mirror glance.

 

[ComUtoR]

They can. ASDO (Automatic Selective Door Operation) does exactly this.

What 'exactly' does ASDO do ?

There are various forms of SDO (selective door operation). The problem is that ASDO or other SDO systems really don't prevent doors opening off a platform.

My TOC uses a GPS based SDO system and a balise based ASDO system. Neither prevent doors opening off a platform.

 

[Bletchleyite]

What 'exactly' does ASDO do ?

ASDO is automatic selective door operation, and it means that when the Open button is pressed, only doors known to be platformed are released. There are several technologies for this, some better than others and some cheaper than others.

GPS is the simplest, for example, but because it can only tell which station it is at it can only release based on the shortest station.

 

[Mojo]

Look back is still used on turbos at certain stations across the GWR DOO network

Also on Chiltern Railways.

 

[ComUtoR]

ASDO is automatic selective door operation, and it means that when the Open button is pressed, only doors known to be platformed are released. There are several technologies for this, some better than others and some cheaper than others.

Apologies, my post was clicked too early so you may have only seen half of it.

Which ASDO system will only release the doors 'known to be platformed' ?

Those who use the system know it's limitations and know where the potential incidents are.

 

[Bletchleyite]

Which ASDO system will only release the doors 'known to be platformed' ?

I believe there are balise based systems that will do that. They're at the expensive end, however.

 

[ComUtoR]

I believe there are balise based systems that will do that. They're at the expensive end, however.

As I stated. We have a balise based system and it does not prevent doors opening off a platform.

 

[Bletchleyite]

As I stated. We have a balise based system and it does not prevent doors opening off a platform.

That doesn't mean there aren't balise based systems that can prevent doors opening off a platform.

 

[ComUtoR]

That doesn't mean there aren't balise based systems that can prevent doors opening off a platform.

I'd agree, hence asking.

Systems like ASDO make for great headlines and are superb at throwaway comments to support a specific side of a debate but when faced with the technical details of what the system does "exactly" then it all gets a little murky.

Please don't think I'm picking on you in any way but you have chosen to engage on a mature level, which is refreshing.

I'm not an enthusiast or professional expert but I would like to understand which systems actually prevent incidents or can genuinely detect doors that are off the platform. Especially if they are on units that are actively in service on DOO routes.

I've used FSDO and that appears to be a decent system (F for 'full') but it's, typically, limited to a single section on the network.

IF DOO came with SDO (A or F) restrictions then I would support it more. It's great saying how awesome DOO is and that ASDO is also just as awesome, but we need to be more honest about the real limitations of the system.

 

[TurboMan]

I don't think there are any ASDO systems on mainline rolling stock that prevent doors being released off a platform in the event of a stop short where the driver releases the doors. There are ASDO systems supplemented by balises (e.g. on 80x which uses Eurobalises, and on Electrostars that use Tracklink 3 balises) but these are all for platform identification, to allow for different door release patterns where the same station has differing platform lengths, and for CSDE (correct-side door enable) to inhibit a wrong-side door release.

I think some London Underground stock uses a balise on the vertical platform face which the balise reader on the train must align with before the doors can be released, but I don't know of any such similar systems on the mainline network. It's a simpler thing to install on the underground as there is generally only one platform in each direction and only one stopping point as all the trains are of the same length.

Having said all that, non-DOO trains run with ASDO, and DOO trains run without ASDO, so the two aren't entirely inter-dependent.

 

[ComUtoR]

I don't think there are any ASDO systems on mainline rolling stock that prevent doors being released off a platform in the event of a stop short where the driver releases the doors.

Thameslink Core does via FSDO

I think some London Underground stock uses a balise on the vertical platform face which the balise reader on the train must align with before the doors can be released, but I don't know of any such similar systems on the mainline network.

I've often wondered why this technology isn't imported to the mainline railway. Is there some kind of limitation ? If an 8@-Car was required to be located at the correct stop point, why would that limit a 10-Car formation ? Is it because you would need multiple beacons ? (Ergo cost?)

It's a simpler thing to install on the underground as there is generally only one platform in each direction and only one stopping point as all the trains are of the same length.

I'll leave that debate to others but Indo understand.

Having said all that, non-DOO trains run with ASDO, and DOO trains run without ASDO, so the two aren't entirely inter-dependent.

Agreed.


Cheers for your post.

 

[HSTEd]

Has anywhere ever attempted remote dispatch via video link from a remote location, for example a control room?

If you are going to have platform cameras, there doesn't seem any inherent reason why the driver has to do it.

You could have someone whos only job is to dispatch trains and can do it all shift without leaving their desk.

 

[Bald Rick]

Has anywhere ever attempted remote dispatch via video link from a remote location, for example a control room?

If you are going to have platform cameras, there doesn't seem any inherent reason why the driver has to do it.

You could have someone whos only job is to dispatch trains and can do it all shift without leaving their desk.

Not in this country, but there’s a little bell going off in my head that a metro system does that somewhere ”forrin“

 

[43066]

ASDO is automatic selective door operation, and it means that when the Open button is pressed, only doors known to be platformed are released. There are several technologies for this, some better than others and some cheaper than others.

The older GPS based systems (SE 375s again as an example, sorry) do this using GPS. This is obviously useful for the many short platforms they serve, but is pretty limited in preventing incidents because it will simply open the correct number of doors assuming the train is correctly platformed. If the driver has stopped short, the same number of doors will release irrespective of still being in a tunnel, on top of an underbridge etc. It will also allow doors to be released on the wrong side, or even stopped at the signal before the station.

I believe there are balise based systems that will do that. They're at the expensive end, however.

As I stated. We have a balise based system and it does not prevent doors opening off a platform.

Has a balise system been installed for 707s to use? I understood they were basically the same as 700s outside the core.

I've often wondered why this technology isn't imported to the mainline railway. Is there some kind of limitation ? If an 8@-Car was required to be located at the correct stop point, why would that limit a 10-Car formation ? Is it because you would need multiple beacons ? (Ergo cost?)

Cost and complexity as always.

Even the GTR FSDO/balise system isn’t perfect and has limits; the related pan up/down system didn’t prevent a train being brought out or Smithfield sidings and heading southbound from deploying its pan not so long ago and blocking the core for several hours.

What would happen if a FLU was stopped at the RLU point? Would all the doors still release? (I know it’s intended that all trains operate though the core on ATO, so shouldn’t happen, but manual driving is still used in certain situations)

 

[Bletchleyite]

Not in this country, but there’s a little bell going off in my head that a metro system does that somewhere ”forrin“

Hamburg's U- and S-Bahn used to sort of do it - there were platform dispatchers in a little glass box, but they mostly did it by looking at screens in front of them, not along the platform, so could theoretically have been anywhere. It's just straight DOO now.

I also wonder what technology might be able to do for this in future. It's not up to it now, and interlock is nowhere near good enough, but if imaging tech can call out a knife in someone's pocket on an airport scanner and highlight it yellow, it can also draw a driver's attention to a potentially obstructed door and yet work out that the other ones are fine. It'd also be able to use far more cameras than a driver can cope with looking at or make sense of, e.g. directly downwards just inside the door.

 

[Mojo]

Has anywhere ever attempted remote dispatch via video link from a remote location, for example a control room?

If you are going to have platform cameras, there doesn't seem any inherent reason why the driver has to do it.

You could have someone whos only job is to dispatch trains and can do it all shift without leaving their desk.

Not in this country, but there’s a little bell going off in my head that a metro system does that somewhere ”forrin“

It is possible on the Victoria and Central lines of the Underground that in the event of defective monitors (ie. the pictures from the platform cameras are not visible in the cab), that the Service Controller can despatch [this is the spelling used by LU] the train using the appropriate cameras in the service control centre.

 

[Bald Rick]

It is possible on the Victoria and Central lines of the Underground that in the event of defective monitors (ie. the pictures from the platform cameras are not visible in the cab), that the Service Controller can despatch [this is the spelling used by LU] the train using the appropriate cameras in the service control centre.

That reminds me - I think this is done (sort of) on SWR at Vauxhall and a couple of the other busier suburban stations - the guard uses a platform monitor to check the further reaches of the train. I’ll leave others to explain why it’s ok for a guard to do it but not a driver.

 

[43066]

That reminds me - I think this is done (sort of) on SWR at Vauxhall and a couple of the other busier suburban stations - the guard uses a platform monitor to check the further reaches of the train.

I suppose that’s still a form of on train dispatch, rather than from a remote location, though.

I’ll leave others to explain why it’s ok for a guard to do it but not a driver.

Now now. If you were driving the train, I guarantee you’d rather the guard did it .

Another slightly random one for @Bletchleyite ’s records, DLR “train captains”, when situated at the front of the train, seem to dispatch using a mixture of platform/tunnel wall mounted monitors and mirrors. Observed today as I used it for part of my journey home due to the strikes!

 

[Bald Rick]

Now now. If you were driving the train, I guarantee you’d rather the guard did it .

Having been up front more times than I care to remember, albeit not driving, I think I’d rather do it myself, preferably with on train monitors.

 

[43066]

Having been up front more times than I care to remember, albeit not driving, I think I’d rather do it myself, preferably with on train monitors.

Well, as you’d likely be an ASLEF member, you might well have voted to agree to do it for a large pay rise .

 

[Bald Rick]

Well, as you’d likely be an ASLEF member, you might well have voted to agree to do it for a large pay rise .

I wouldn’t be

 

[ModernRailways]

That reminds me - I think this is done (sort of) on SWR at Vauxhall and a couple of the other busier suburban stations - the guard uses a platform monitor to check the further reaches of the train. I’ll leave others to explain why it’s ok for a guard to do it but not a driver.

This is also how it is done in Japan, guard visibly checks what they can see but uses cameras to view the full length of the train, particularly at the peak times where platforms are crowded so seeing the full length of the train may be difficult. The guard is (should be?) at the back of the train, so if they did end up with a trap and drag they are in a position to see that and to stop the train once it's moving. If it was driver only would this be possible once the train is moving? Or once power is applied would a driver be focused fully back on driving the train and not looking at in cab cctv, platform monitors etc?

 

[JamesT]

This is also how it is done in Japan, guard visibly checks what they can see but uses cameras to view the full length of the train, particularly at the peak times where platforms are crowded so seeing the full length of the train may be difficult. The guard is (should be?) at the back of the train, so if they did end up with a trap and drag they are in a position to see that and to stop the train once it's moving. If it was driver only would this be possible once the train is moving? Or once power is applied would a driver be focused fully back on driving the train and not looking at in cab cctv, platform monitors etc?

Is the guard going to be any better at seeing a trap and drag? Pretty much by definition the person is being carried along so won’t end up at the back of the train. There’s probably no drop light for them to be looking along the train as it sets off.

 

[ModernRailways]

Is the guard going to be any better at seeing a trap and drag? Pretty much by definition the person is being carried along so won’t end up at the back of the train. There’s probably no drop light for them to be looking along the train as it sets off.

They would be in a position to see people or platform staff on the platform reacting and thus be in a position to react much as they would if there were no cameras on the platform. I’m not 100% sure on the role of the guard, but from what I know, the guard remains at their door, or window, and looks out to ensure the safe departure of the train. If someone is trapped then they would hopefully see this before it happens and the train departs but there’s always extremes where it can’t be seen (doors closed, then someone puts hand in to try and pry doors open, for example) and the guard would be able to react to such a situation, whether it be from them seeing it, or because other passengers/platform staff have alerted them to it. Would a driver only be able to react in the same way considering their main task is to focus on the line ahead, and they can’t exactly watch a 12 car train leave the platform or that would introduce a huge spad risk?

 

[philthetube]

Also on Chiltern Railways.

YOu see them doing it at the MET stations and they are not supposed to

I think some London Underground stock uses a balise on the vertical platform face which the balise reader on the train must align with before the doors can be released, but I don't know of any such similar systems on the mainline network. It's a simpler thing to install on the underground as there is generally only one platform in each direction and only one stopping point as all the trains are of the same length.

This is correct at least on S stock

Additionally before the driver can open the doors if either the equipment fails or if away from a platform the driver must firstly open a locked cabinet operate a switch, then open the cab door on the platform side and only then the doors can be opened. Drivers are also trained to put a foot on the platform before opening the doors, this is all to prevent wrong side door opening.

It is possible on the Victoria and Central lines of the Underground that in the event of defective monitors (ie. the pictures from the platform cameras are not visible in the cab), that the Service Controller can despatch [this is the spelling used by LU] the train using the appropriate cameras in the service control centre.

But not on the MET, the cameras which control access are not the same ones as drivers use and are focused on the platform in general and not the train doors.

I must admit I like the platform cameras on train monitors, I also like being able to see the platforms while departing,

I agree that this is not possible on the main line so don't bite my head off).

The underground is a closed system so there should not be anything going on ahead of the train, except a station starter, (Signal). So the driver has less to worry about, drivers also have a more generous spad allowance.

If you are wary of something going on on the platform, kids, drunks or whatever, it is possible to depart slowly and be ready to stop if needed.

I have seen train borne camera images and don't find them half as clear, and there are more images to look at, platform cameras provide a max of 8 on a 8 car train is 8, straight platforms have two.

 

[ComUtoR]

Has a balise system been installed for 707s to use? I understood they were basically the same as 700s outside the core.

Tracklink 3 I believe. It's a very different system.

Cost and complexity as always.

Which is so disappointing. PTI incidents can be designed out with minimal cost.

What would happen if a FLU was stopped at the RLU point? Would all the doors still release? (I know it’s intended that all trains operate though the core on ATO, so shouldn’t happen, but manual driving is still used in certain situations)

Never driven an FLU so I can't answer, sorry.

Having been up front more times than I care to remember, albeit not driving, I think I’d rather do it myself, preferably with on train monitors.

It's different when you have to press the button. Doing it once or twice may be a novelty and quite interesting, doing it 100 times a day is a lot different. Doing it in the peak can vary from frustrating to absolutely scary.

 

[TurboMan]

Would a driver only be able to react in the same way considering their main task is to focus on the line ahead, and they can’t exactly watch a 12 car train leave the platform or that would introduce a huge spad risk?

From RSSB's T1059 research project, which evaluated the risks associated with drivers using in-cab monitors while starting away:

...the conclusion is that there is no distinct difference, suggesting that leaving the monitors on does not impact on overall system risk when compared with the monitors being off. The monitors remaining on provides additional benefits in managing risk at the PTI by allowing drivers to identify late presenting hazards or rectify errors made in initial checks of the monitors.
However, the requirement to share attention between tasks needs to be considered, in particular at locations where the driver will be required to focus their attention on the track ahead due to location or time specific hazards. This can be managed through local instructions to the driver, driver competence and measures, such as TPWS, to minimise the consequences of driver error.

The full research is here (but you'll need an RSSB account to access it): https://www.rssb.co.uk/en/research-catalogue/CatalogueItem/T1059

== Doublepost prevention - post automatically merged: 14 Dec 2022 ==

Thameslink Core does via FSDO

I didn't know that, I thought it was automatic.

I've often wondered why this technology isn't imported to the mainline railway. Is there some kind of limitation ? If an 8@-Car was required to be located at the correct stop point, why would that limit a 10-Car formation ? Is it because you would need multiple beacons ? (Ergo cost?)

I imagine so. The beacons would need to be coded for the train length at the PSM, and the train would need to be able to read the beacon and detect that it is at the correct one, otherwise there is nothing to prevent doors being released on a 10-car stopping short at the 8-car PSM, and detecting the balise there.

 

[Bletchleyite]

Watching the train out is also something artificial intelligence image analysis could potentially help with. Something moving precisely with the train that shouldn't be is not a very difficult thing to identify.

DOO or no, I have long considered it a major gap; the guard can't watch the train out on most units. And you're more likely to see a trap and drag once moving as people react to the incident.

 

[Horizon22]

First of all I should state that this thread is NOT a pro/anti DOO debate and not a place for related antagonism/arguments. I'd ask the mods, if they would, to delete any posts which take it this way. This is about existing DOO routes.

A question for drivers of DOO or anyone else with an interest - what method(s) of DOO dispatch do you prefer and in what contexts?

I guess the options are:
1. Dispatchers and CD/RA
2. Convex mirrors
3. Platform based cameras with on-train display
4. Platform based cameras with screens on the platform (used to be common on London Underground)
5. On-train cameras
6. On-tram wing mirrors (don't think there are any UK heavy rail applications)

I suspect most will prefer 1, but what of the others? Obviously practicality varies, e.g. convex mirrors aren't going to be of use for a 12-car behemoth on a curved platform.

1. Probably the best method, because there is an opportunity for a visible staff presence too. Is it viable for tiny halts with low passenger numbers though? Also the infrastructure needed to install CD/RA everywhere is probably more costly than just utilising the train infreastructure - much new stock has this installed by default.
2. Only really viable I think for extremely straight platforms and even then, I don't think it should be acceptable for anything longer than a 5-car. But it is in place for longer trains!
3&5 are somewhat linked in that they both have an on-train display. Only drivers can commentate on this day-in-day-out but the vast majority of the time cameras are very clear.
6. Not sure that's a thing as you state and sounds like an ineffective way.

You missed option 7 - look back (although I suspect it will not be warmly favoured!)

It's available as a degraded method I think at some TOCs.

 

[ComUtoR]

Watching the train out is also something artificial intelligence image analysis could potentially help with. Something moving precisely with the train that shouldn't be is not a very difficult thing to identify.

It's superfluous. The train safety check takes place BEFORE the train moves. Once the train is moving then all bets are off.

DOO or no, I have long considered it a major gap; the guard can't watch the train out on most units. And you're more likely to see a trap and drag once moving as people react to the incident.

With DOO, once the unit moves then it's in the hands of the Gods. At least with a Guard or Platform staff you have a chance.