Train Driving and Automation

[Dave1987]

I think (keeping things on topic) that part of the problems with getting full ato onto the mainline is precisely because we see the smaller forms of automation, such as the sander, still not being able to properly cope with the reality of train driving. ATO would have a higher degree of safety protocols etc but it still clearly has a very very long way to go.

If we can't get the sander right then how are we to get a full blown automated train right.

NR can't even get simple Track circuit block signalling to work reliably let alone the infrastructure required for full blown automation!

 

[D1009]

It's hard to quantify because as part of the upgrade the traction power of the trains was increased (previously the motor output was capped by software restriction), and also maximum line speed has been increased from 40 or 45 mph to 50 mph.

What I was really after was what effect the introduction of ATO has had on the working timetable on the Northern or Central Lines.

 

[DerekC]

Yes. The reason given is because of the likelihood of the VOBCs halting if the train goes into a slide, and the time then taken for the VOBCs to be re-entered into the system, which involves travelling at slow speed to the next inductive loop. In some cases the gentle brake rate will be all the way down from 50 mph, and in the future may be from 62 mph which is intended to be the maximum speed in the long-term on selected sections

Interesting - that's precisely the same issue as DLR had in about 1995. DLR had two tachometers and an accelerometer feeding each VOBC, the idea being that the accelerometer can fill in the gaps when the wheels side for a short period. We did get it to work fairly reliably after a lot of parameter tweaking and some software mods. I thought SELTRAC had got a bit more sophisticated, but it sounds like not. The basic problem is that it was designed originally for the Vanncouver Skytrain which has linear motors and unbraked axles (or at least ones where the brakes only apply in an emergency). ATO over ETCS has a chance of being better because the ETCS uses two tachos plus a Doppler radar, which should give plausible position for longer during a slide.

 

[Goldfish62]

What I was really after was what effect the introduction of ATO has had on the working timetable on the Northern or Central Lines.

That's easy to answer. It has increased capacity. There wouldn't have been any point in introducing it otherwise. I remember when ATO was introduced on the Central line some drivers were alarmed at how fast their trains were now entering platforms.

I know there are some issues with set speeds and braking points on the Northern at the moment, but overall the increased speed on the Charing X branch at least is noticeable.

 

[c52]

What are the prospects for making trains automatically and gently couple and uncouple in the amount of time a train normally spends in a platform?

 

[bramling]

That's easy to answer. It has increased capacity. There wouldn't have been any point in introducing it otherwise. I remember when ATO was introduced on the Central line some drivers were alarmed at how fast their trains were now entering platforms.

I know there are some issues with set speeds and braking points on the Northern at the moment, but overall the increased speed on the Charing X branch at least is noticeable.

It's not as clear cut as this. The Northern Line is currently not delivering anything more than a conventionally signalled line could deliver. Personally I have doubts that the Northern Line service will be able to be enhanced much above what the current timetable delivers. There are a number of pinch points with the new signalling where Seltrac struggles to maintain the throughput of the old signalling. From a driver's perspective, if there's a train in the platform ahead, Seltrac will bring the train nearer to the platform, but then it will wait for a long time before getting going again, in many cases long after all three home signals would have cleared under the old signalling. One of the worst places is Stockwell southbound, which just happens to be one of the biggest bottlenecks on the entire line in the evening peak! Likewise approaching termini, the train gets held way further back than where the old home signals used to be.

To be fair, the Victoria Line system is much better.

 

[D1009]

That's easy to answer. It has increased capacity. There wouldn't have been any point in introducing it otherwise. I remember when ATO was introduced on the Central line some drivers were alarmed at how fast their trains were now entering platforms.

I know there are some issues with set speeds and braking points on the Northern at the moment, but overall the increased speed on the Charing X branch at least is noticeable.

Sorry, should have made myself clearer. I meant what effect has the introduction of ATO had on scheduled journey times on the surface sections of the Central or Northern lines.

 

[Chris M]

That's easy to answer. It has increased capacity. There wouldn't have been any point in introducing it otherwise.

To me this is the key point about why we haven't seen a wider rollout of automated driving on the railways - outside of metro environments there just isn't the need for capacity above what conventional operation can deliver.

Introducing ATO based on today's technology requires very extensive (and thus expensive) programming of all details of speeds, gradients, stopping points, etc. I wouldn't be surprised if the expense increased faster than the distance either.

There are undoubtedly issues with existing system, but given enough time and money these are all resolvable. However, there is not the will currently to spend the money, as the benefits are not proportionate in most railway environments.

 

[Bromley boy]

Fascinating thread which I have read with interest.

This is a subject I'm interested in as a new entrant to the driver grade with 30 years of my working life left. I even discussed This topic in my DM interview.

What has reassured me about my own shelf life is that, although the technology to make driverless trains a reality is clearly already present (and has been for many years), the wider rail industry moves at a glacial pace. Some topics on the current drivers rules course can be demonstrated using videos from as far back as the 1950s, and further. Not because of nostalgia, but because this is how large swathes of the UK network still operates in 2015.

The real question is whether the will is there to spend the necessary cash to deliver full automation. The network is largely Victorian and requires massive amounts of investment just to maintain it at operational standard. The level of investment made to introduce full automatic operation is astronomical. It is also of questionable benefit when we already have the safest railway in Europe.

GSM-R is a "new" technology to the railway but actually relies on early 90s 2G technology. ERTMS is gradually being rolled out but isn't expected in the region I will be driving in until around 2040, by which time I will be approaching retirement. Even then the level 2 standard being adopted will still require a driver at the controls.

As with the aviation industry I reckon the real point is, although technology may be available for pure automation, will the public ever accept it? As noted in previous posts the DLR has been driverless since its inception, but a member of staff is carried on every train and these staff are still capable of bringing the railway to a grinding halt when they strike.

All in all I am confident that I will have a valued job as a driver for the remainder of my working life - albeit perhaps switching from a "driving" towards a "monitoring" role as the years past.

I sincerely hope I'm right!

 

[DerekC]

What has reassured me about my own shelf life is that, although the technology to make driverless trains a reality is clearly already present (and has been for many years), the wider rail industry moves at a glacial pace

.

Absolutely right about the glacial pace.

The real question is whether the will is there to spend the necessary cash to deliver full automation. The network is largely Victorian and requires massive amounts of investment just to maintain it at operational standard. The level of investment made to introduce full automatic operation is astronomical. It is also of questionable benefit when we already have the safest railway in Europe.

Depends what you mean by "full automation". If "unattended trains" then I agree, basically because the railway simply isn't anywhere near reliable enough to support this. However once ERTMS is in place then the extra investment for an "autopilot for train drivers" is relatively small when you are buying new trains. Modifying old ones (especially very old ones) can cost a lot.

GSM-R is a "new" technology to the railway but actually relies on early 90s 2G technology. ERTMS is gradually being rolled out but isn't expected in the region I will be driving in until around 2040, by which time I will be approaching retirement. Even then the level 2 standard being adopted will still require a driver at the controls.

Yes but see above regarding the extra costs - the autopilot can be added on to Level 2.

As with the aviation industry I reckon the real point is, although technology may be available for pure automation, will the public ever accept it? As noted in previous posts the DLR has been driverless since its inception, but a member of staff is carried on every train and these staff are still capable of bringing the railway to a grinding halt when they strike.

The public likes DLR for two reasons - first because their kids can sit in the front and pretend to be a train driver (probably not transferable to the main line) and second because there is a member of train crew in a public facing role fairly near them (used to be even better when the trains only had one unit). I don't think having a train driven automatically will worry most people. The point is that there is a railway staff person on the train.

All in all I am confident that I will have a valued job as a driver for the remainder of my working life - albeit perhaps switching from a "driving" towards a "monitoring" role as the years past. I sincerely hope I'm right!

I think you are right - but what do you want your "monitoring" role to be - and how will you stay awake during a 2 - 3 hour journey for the occasional few minutes when you are needed?

 

[Bromley boy]

.

Absolutely right about the glacial pace.



Depends what you mean by "full automation". If "unattended trains" then I agree, basically because the railway simply isn't anywhere near reliable enough to support this. However once ERTMS is in place then the extra investment for an "autopilot for train drivers" is relatively small when you are buying new trains. Modifying old ones (especially very old ones) can cost a lot.



Yes but see above regarding the extra costs - the autopilot can be added on to Level 2.



The public likes DLR for two reasons - first because their kids can sit in the front and pretend to be a train driver (probably not transferable to the main line) and second because there is a member of train crew in a public facing role fairly near them (used to be even better when the trains only had one unit). I don't think having a train driven automatically will worry most people. The point is that there is a railway staff person on the train.



I think you are right - but what do you want your "monitoring" role to be - and how will you stay awake during a 2 - 3 hour journey for the occasional few minutes when you are needed?

All very true. Whilst I don't doubt the virtues of ERTMS/ATO from an efficiency and capacity perspective, I'm hoping for a "driving" role for a good few years yet. Nothing I've seen or heard has indicated this shouldn't be the case.

From a purely selfish point of view this is very important given I've changed a well paid career for train driving which, although also well paid, has limited transferable skills...