Train Driving and Automation

[DerekC]

Many will already know that Thameslink is pioneering the application of ATO (Automatic Train Operation) on the main line railway. This is being done using ERTMS as the ATP safety underlay for ATO. Assuming it works, is this the thin end of a very large wedge? Will ATO be overlaid on all ERTMS routes, and if so what will the main line train driver's role be in the future?

 

[PeterY]

I've often thought about automatic train operation. If I were a driver, for arguments sake on the Central line in London, I think I be bored stiff and I'm sure my mind would start thinking about other things.

 

[Aivilo]

It won't happen in my career so not overly fussed

 

[Bald Rick]

Very likely, progressive introduction over next 20-30 years to all major routes.

 

[Aivilo]

We run on semaphores and non electrified tracks. Think it'll be a dam sight longer

 

[D1009]

This has already be done almost to death in previous threads. So far it has only worked properly on underground railways, and it's extension to surface parts of the LU network has caused problems.

 

[Carlisle]

Very likely, progressive introduction over next 20-30 years to all major routes.

It depends on what technology comes on stream and how easy and affordable it is to buy and install on existing routes, so right now I'd say no one really knows

 

[ComUtoR]

I'm looking forward to it.

 

[Bald Rick]

This has already be done almost to death in previous threads. So far it has only worked properly on underground railways, and it's extension to surface parts of the LU network has caused problems.

Really? Seems to have worked fine on the DLR for nearly 30 years. And the Central line for 25.
--- old post above --- --- new post below ---

It depends on what technology comes on stream and how easy and affordable it is to buy and install on existing routes, so right now I'd say no one really knows

The technology is called ETCS, with ATO overlay. The former is here now and works on existing routes. The latter is under test and will be working next year (albeit not in passenger service for a year or two longer).

Cost is a factor of course, but I would be astonished if every new fleet of trains ordered for major commuter routes after 2018 isn't equipped with this kit as standard. Certainly HS2 trains will be.

 

[bramling]

Really? Seems to have worked fine on the DLR for nearly 30 years. And the Central line for 25.
--- old post above --- --- new post below ---


The technology is called ETCS, with ATO overlay. The former is here now and works on existing routes. The latter is under test and will be working next year (albeit not in passenger service for a year or two longer).

Cost is a factor of course, but I would be astonished if every new fleet of trains ordered for major commuter routes after 2018 isn't equipped with this kit as standard. Certainly HS2 trains will be.

I wouldn't agree that ATO on the surface parts of LU has gone well.

In the relatively tame environment, the difficulty of coping with different and changing rail conditions in the open has not been fully resolved. All of the railways, including DLR, have measures for mitigating the issue, which impose constraints on operation. These constraints and mitigations just about work on a metro railway. As speed increases these issues tend to become more of a problem.

Until we develop a reliable method of identifying changing railhead conditions, ideally before they occur, I can't see widespread ATO coming to the mainline. It's not that it can't be done, but more that at present the drawbacks probably outweigh any potential benefits.

 

[HLE]

Many will already know that Thameslink is pioneering the application of ATO (Automatic Train Operation) on the main line railway. This is being done using ERTMS as the ATP safety underlay for ATO. Assuming it works, is this the thin end of a very large wedge? Will ATO be overlaid on all ERTMS routes, and if so what will the main line train driver's role be in the future?

What makes me laugh is that with a growing population why are we trying to automate somebody's job?

Surely we need more jobs not less!

 

[Cherry_Picker]

Microeconomics vs macroeconomics is the reason why. Companies/industries will do what's right for them, not what's right for the overall economy. It might turn out to be a disaster, it might turn out to be okay. 150 years ago most of the world was an agricultural society then the industrial revolution happened and people at the time were up in arms about how machines were taking peoples jobs (organised groups like the Luddites were actually organised to combat it) but I don't think many people in developed countries would go back to how it was before.

I'm not saying I want automation to come because it will put me out of a job but I accept that it's inevitable. From a selfish point of view I hope my career ends in retirement before I'm made obsolete by technology and given the cost of upgrading infrastructure on existing lines then maybe I'll make it out unschathed. I'd certainly be a lot more worried about automation having a direct impact on my career if I were a HGV driver than a train driver.

 

[Dave1987]

As has been said in many many threads before now, it will cost billions and billions of £'s to implement and you would still need a fully qualified driver on board so I'm not worried even slightly that my job will be automated. The UK is renowned for running about 20-30 yrs behind technology and the rest of the world.

 

[dk1]

It won't happen in my career so not overly fussed

Mine neither so not fussed at all.

 

[DerekC]

Thanks for some interesting stuff already. As a newbie I didn't know there was lots of stuff in previous threads. Combining the inputs from Bald Rick, Bramling and Dave 1987 I get:

1) The technology as such isn't the problem (ETCS plus ATO should work, although the GSM-R comms is still a weak link)

2) Adhesion and railhead condition is a key issue, but LU has overcome it on its surface sections with a combination of local detailed weather forecasting and prediction, feedback from WSP on the trains and reducing the ATO brake rate on the basis of predicted conditions. They have a special software package in the control room for bringing it all together. Something similar should work on the main line , at least in the sections where this technology is worth applying (let's think SWML, for example).

3) But - and it's a big but - the main line railway is nowhere near reliable enough (probably by a factor of 10) to run trains in unattended mode, or even DLR style with a steward to take over when something breaks - think about a getting to the cab in a crush-loaded 12 car train. So we will need someone in the cab for a long time to come.

My question is - assuming it happens what will the driving job be like on the likely automated main line railway of 20 years time and how do we make that work on a human basis, balancing the need for alertness against boredom?

 

[notadriver]

Point 2 - I've ridden in the cab of an ATO train and it braked very cautiously on the surface sections. I've seen central line trains stop short and need manual intervention to pull up to the correct stopping point. So the the problem is far from being overcome. Have you driven a train in low adhesion ? I did as recently as yesterday and you need to do that to understand the problem.

 

[HLE]

If you are going to have to have a qualified driver on board anyway with automation why not just let him/her drive.

Such a waste of money, for very small gains

 

[bramling]

Thanks for some interesting stuff already. As a newbie I didn't know there was lots of stuff in previous threads. Combining the inputs from Bald Rick, Bramling and Dave 1987 I get:

1) The technology as such isn't the problem (ETCS plus ATO should work, although the GSM-R comms is still a weak link)

2) Adhesion and railhead condition is a key issue, but LU has overcome it on its surface sections with a combination of local detailed weather forecasting and prediction, feedback from WSP on the trains and reducing the ATO brake rate on the basis of predicted conditions. They have a special software package in the control room for bringing it all together. Something similar should work on the main line , at least in the sections where this technology is worth applying (let's think SWML, for example).

3) But - and it's a big but - the main line railway is nowhere near reliable enough (probably by a factor of 10) to run trains in unattended mode, or even DLR style with a steward to take over when something breaks - think about a getting to the cab in a crush-loaded 12 car train. So we will need someone in the cab for a long time to come.

My question is - assuming it happens what will the driving job be like on the likely automated main line railway of 20 years time and how do we make that work on a human basis, balancing the need for alertness against boredom?

LU hasn't really solved the problem - they have put in place enough mitigations so that it just about works.

- The Central Line runs sandite trains in the autumn, and tolerates a level of overshoots and operating incidents. Manual driving is frequently used to get out of trouble.

- The Jubilee and Northern Lines uses the most gentle braking rate available in the open. This mostly avoids incidents, but at a massive cost in terms of extra running time. Fortunately this only affects a proportion of the railway.

- The Victoria Line is all in tunnel so this doesn't apply.

- DLR vary their brake rate daily according to forecast. Again this imposes slow running when not required. DLR is not badly affected by changing conditions as it doesn't suffer badly fron leaf-fall, and is a metro railway with high frequency which also helps.

ATO has benefits on high-frequency tunnel metro railways, but elsewhere I don't believe the benefits outweigh the costs. And don't believe ATO is all it's cracked up to be, on the Northern Line it's quite easy to drive faster and make up more time in Protected Manual rather than ATO due to some of the strange things the ATO does which loses time here and there, again all the LU systems do this in places. Anyone who thinks ATO is perfect and wonderful has never seen it in action.

 

[Dave1987]

What we need to take into consideration here is that the Govt treats things like this in terms of Cost - Benefit ratio. So how much benefit would you get from spending all that money on putting in ATO. It's all well and good using adhesion forecasts to vary the braking on your automated train but in the real world it doesn't world it doesn't work like that. Any driver who has driven a route for a decent amount of time will have the experience to read the conditions at that time and know whether they need to brake early or not. It think widespread DOO is inevitable regardless of how much we would rather it didn't come in. But the costs of ATO really aren't worth it considering that you would still need a driver on the train. ERTMS will be rolled out over the next decade or so but that does provide a real improvement in track capacity although I'm sceptical about how reliable it will be considering how we were told GSMR would never lose signal

 

[al78]

What makes me laugh is that with a growing population why are we trying to automate somebody's job?

Surely we need more jobs not less!

The problem with that argument is that you could extend it to everything where computers and machines are used, and so any sort of automation or technology which might negate the services of a human is a bad thing. Perhaps if we are really at that stage in regard to job availability we should give up private cars and all use taxis or public transport instead, or pay cleaners for our houses instead of doing it ourselves, or pay tradespeople instead of doing DIY, or go shopping on the high street instead of on the internet to help prevent the closure of town centre shops. Where do you draw the line?

For something like automation of driving vehicles there will be an initial research and development phase which will require the employment of scientists and engineers for a lengthy period (I don't anticipate automation of trains to be a trivial thing to implement), so at least initially there will be people employed to look at the problem and innovate, as is the case for any advancement in technology. But in the end it will come down to a cost benefit analysis as to whether it is worth implementing.

 

[bramling]

The problem with that argument is that you could extend it to everything where computers and machines are used, and so any sort of automation or technology which might negate the services of a human is a bad thing. Perhaps if we are really at that stage in regard to job availability we should give up private cars and all use taxis or public transport instead, or pay cleaners for our houses instead of doing it ourselves, or pay tradespeople instead of doing DIY, or go shopping on the high street instead of on the internet to help prevent the closure of town centre shops. Where do you draw the line?

For something like automation of driving vehicles there will be an initial research and development phase which will require the employment of scientists and engineers for a lengthy period (I don't anticipate automation of trains to be a trivial thing to implement), so at least initially there will be people employed to look at the problem and innovate, as is the case for any advancement in technology. But in the end it will come down to a cost benefit analysis as to whether it is worth implementing.

Agree with all of this. For those who seem to think automation is the utopian answer to everything, this is simply delusional.

The benefits of ATO arise in a metro environment when you can achieve a more consistent and faster driving style.

On the mainline, at present the more consistent driving style will probably mean a 'leaf fall' timetable all the time. Safety performance on the mainline is nowadays excellent and I don't see any political desire to 'do something', in any event ATP provides all the benefits without the drawbacks. Finally you will need a qualified individual on the front of the train, so you don't get the benefits from being able to abandon the bane of coverage/duty schedules etc.

At present, for most applications, having a competent skilled and well trained driver better and more cost-effective than ATO. We've already had ATO for over 50 years and today's technology has gone little further into designing out the problem of railhead variability.

In summary, we are nowhere near having a system suitable for nation-wide ATO, that can maintain let alone improve upon current driver performance. Thameslink and Crossrail are buy no means typical applications, neither is LU.

 

[DerekC]

There is a strong thread here of not wanting to think about something which is almost certainly going to happen. It is odd how many people in the railway industry believe they will have an automated car in 20 years time, but trains will still be manually driven!

The benefits of ATO are consistent performance in speed control, reduced platform reoccupation times, shorter dwell times (because the train has to know where it is when it stops, so doors can open immediately) and better PTI safety on departure (because the driver doesn't have to worry about controlling the train). Also improved energy efficiency and immediate response to regulation instructions. And the one that nobody likes to mention - greatly reduced dependence on route knowledge, so greater flexibility in rostering and reduced training need.

The problem currently is that nobody is thinking about this from the driver's perspective. The engineers are assuming that metro technology and operational concepts will do for main lines (which I agree they won't) and the drivers don't want to think about it.

If ATO over ETCS on Thameslink is successful (we should know in 3 years time) and ETCS gets rolled out on ECML south by 2020-ish, the costs of extending ATO up ECML for the Class 700s will be fairly modest, so things might start to happen. The Class 700s also happen to be long (min 8 cars) and fitted with modern brakes combining clasp (motor bogies) and disc (trailers) so should be as good as it gets in terms of stopping in poor adhesion conditions.

 

[dcsprior]

What makes me laugh is that with a growing population why are we trying to automate somebody's job?

Surely we need more jobs not less!

I think you're either 100% right or 100% wrong

When I was a child, one of my favourite programmes on the TV was Tommorow's World. I fully expected that by time I enterred the world of work, no-one would be working more than a 3 day week.

Over sixteen years after entering the world of full-time work, I'm still waiting for this! In fact, the reverse has been true as like many people I've had to work extra, initially because I needed the paid overtime to make ends meet; and more recently sometimes doing extra (unpaid) hours to ensure I can get through my workload. At the same time, the age at which is considered a "normal" retirement age for pensions purposes keeps rising. Maybe this is why people refer to a generation of people who feel like the world owes them a living - I does feel like we were promised better than this.

So, if more automation happens across multiple industries and there's a concerted effort to ensure that this means that there are still jobs for everyone with each individual working less and retiring at a decent age, then I think you're 100% wrong.

However if it we continue to organise our society along the "live to work" lines that we have at present and automation just results in the people previously doing a well-paid job being thrown on the dole, then I think you're 100% right.

Unfortunately, I think its likely you are 100% right

 

[D1009]

It's not going to be worth investing money in developing ATO on the main line if it results in journey times being increased or the service becoming more unreliable in bad weather.

 

[najaB]

There is a strong thread here of not wanting to think about something which is almost certainly going to happen.

I think the common thread is that while it technically is possible, it is doubtful that it will be worth it. All the benefits you outlined can be added as driver aids without moving to full ATO.

 

[bramling]

There is a strong thread here of not wanting to think about something which is almost certainly going to happen. It is odd how many people in the railway industry believe they will have an automated car in 20 years time, but trains will still be manually driven!

The benefits of ATO are consistent performance in speed control, reduced platform reoccupation times, shorter dwell times (because the train has to know where it is when it stops, so doors can open immediately) and better PTI safety on departure (because the driver doesn't have to worry about controlling the train). Also improved energy efficiency and immediate response to regulation instructions. And the one that nobody likes to mention - greatly reduced dependence on route knowledge, so greater flexibility in rostering and reduced training need.

The problem currently is that nobody is thinking about this from the driver's perspective. The engineers are assuming that metro technology and operational concepts will do for main lines (which I agree they won't) and the drivers don't want to think about it.

If ATO over ETCS on Thameslink is successful (we should know in 3 years time) and ETCS gets rolled out on ECML south by 2020-ish, the costs of extending ATO up ECML for the Class 700s will be fairly modest, so things might start to happen. The Class 700s also happen to be long (min 8 cars) and fitted with modern brakes combining clasp (motor bogies) and disc (trailers) so should be as good as it gets in terms of stopping in poor adhesion conditions.

I don't envisage having an automtically driven car, and would feel very uncomfortable with the concept. No automatic system could decide not to drive down an iced-over hill, for example.

All the benefits of ATO you describe pertain to metro applications. The more you move away from that application, the less benefit you get, and the more issues start to arise.

It's not that noone has thought about it, on the contrary having considered the benefits and costs, the tipping point currently lies at the point where you transition from metro to mainline. A network like the Kent Link suburban lines, or SWT suburban, may lie on the tipping point, but even there I don't think the technology favours ATO. The modern railway and professional driving policies have introduced greater consistency to driving anyway. All the other so-called benefits, eg quicker door release, are tiny compared to the potential for lost time on good railhead conditions. Meanwhile, again looking at the Northern Line, notice how there is a time lag between train stopping and doors releasing - again don't believe all you hear about how wonderful ATO is.

 

[notadriver]

I'm currently on a Central line train. It's come to a very abrupt stop halfway down the platform (again) [emoji28][emoji28][emoji28]

 

[DerekC]

However if it we continue to organise our society along the "live to work" lines that we have at present and automation just results in the people previously doing a well-paid job being thrown on the dole, then I think you're 100% right.

Unfortunately, I think its likely you are 100% right

I wouldn't want to claim to be 100% right about anything, but I am afraid that is exactly what is happening. Automation results in more profits for those at the top of the financial tree and lower paid jobs and less autonomy and interest for those at the bottom. It has been like that since about 1780 or so! What seems to be getting worse is that those at the top now think it's quite OK to avoid paying the taxes which should go to evening things up a bit.

What worries me in the context of automated driving is that so many drivers are kidding themselves that it isn't going to happen. I don't want to get into an "I know better than you" exchange with anyone who has contributed to this thread, but I have spent a good chunk of the last 25 years commissioning ATO systems so I do know they aren't easy to get right - but they can be got right, and the incentlve to get them right for main lines is there. It will start with high density railways - in fact it has already started - and move out from there. I saw in another thread people blithely assuming that in future there would be 6 tracks from New Malden to Clapham Junction (in addition to Xrail2). It ain't going to happen - the cost of one extra track has already been estimated as £5Bn, quite apart from the "protect the back gardens of South London" lobby which will promptly spring up. The pressure will be on to get a lot more trains over the tracks we have. And yes, I know it's a lot to do with service mix and stopping pattern, but part of the solution will be ATO.

Anyhow, this thread doesn't seem to be going the way I hoped (i.e. a discussion about the future role of the main line driver). One last time - are there some thoughts about that out there?

 

[notadriver]

Every so often someone - usually a new member like yourself DerekC, brings up the subject of automation/driverless trains etc. They then don't like the responses being given and the whole thing descends into a 'I know best contest' . You've heard the responses from drivers and others, that isn't going to change so let's just leave at that - yes ? [emoji4]

 

[Cherry_Picker]

There is a strong thread here of not wanting to think about something which is almost certainly going to happen. It is odd how many people in the railway industry believe they will have an automated car in 20 years time, but trains will still be manually driven!

The reason for that is probably that there is significantly more research and development into automatic cars. Maybe it's an easier problem to solve as all the technology needed is contained within the car and currently the best solutions we have for heavy rail is an interface between train and track that doesn't scale very well over long distances. Self driving cars are already on the roads and the financial motivation to push on with them is huge, probably far bigger than the rail network in fact. The cost of transporting freight via road would plummet overnight and vehicles could go much further in a 24 hour period as they wouldn't be slave to the dreaded tachograph.

Longer term, 50-100 years if automated cars really take off then a far smaller percentage of the population will own a car, especially in urban areas. Almost everything will become a taxi and it will reduce the need for car parking space in central areas of major towns and cities. Imagine if all the car parking space in zone 1 of london could be repurposed for something else?