Public's misconceptions of the railway..Driving is easy... funny and aggravating..

[DD]

I think if you spent a week on the ground you would see the amount of issues that arise that without a driver would cause massive disruption to the service. anyhow ertms will take over 50years to implement and they are still having issues with the trial in this country with arriva saying that service is at an all time low. Drivers will be present for years to come.

Ill reillustrate notadrivers post from above.

 

[cjp]

And Japan also has a history of staff taking risks that costs lives in an attempt to maintain that efficiency.
See the Amagasaki accident for an example where the Japanese obsession with full manual control of trains lead to numerous deaths.

A computer controlled train cannot be driven to break speed limits and take other unsafe actions in an attempt to keep to time.
The only reason to retain drivers at the present time is for collision avoidance, more modern lines will likely be fitted with PEDs universally in the future which further reduces the need for them to be retained.

The fact that in most cases this is assumed to be a driver is irrelevant, the system will work just as well with a computer in control assuming you have some mechanism to detect obstructions on the line ahead or have taken steps to exclude the possibility of such things.

A driver seeing an obstruction at speed will not permit them sufficient time to brake to stop before impact with the obstruction.
All they might have is time to make an emergency brake application and to shut their eyes.

The days of the railway children and waving a red cloth are long gone.:cry:

 

[HSTEd]

A driver seeing an obstruction at speed will not permit them sufficient time to brake to stop before impact with the obstruction.
All they might have is time to make an emergency brake application and to shut their eyes.

The days of the railway children and waving a red cloth are long gone.:cry:

Hitting an obstruction at 100mph instead of 125mph is likely to reduce damage considerably though.

It goes from being Eschede to an accident like Great Heck.

 

[Tomnick]

As 'DD' suggests above, I think there's a lack of understanding here of some of the things that Drivers have to deal with on a daily basis. It's not just about ploughing on and dropping the lot when you see an obstruction - Drivers report all sorts of things, from rough rides (could indicate all sorts from a broken rail to a collapsed bridge) and UWC gates left open to an obstruction blocking the opposite line, so that they can then be sorted as necessary. You've then got the issue of examining the line, or running under caution, which is admittedly more readily dealt with once everything's been brought to a stand, but still needs considering. It's less of a problem on high speed lines with no level crossings, secure fencing and impressive engineering, but it's the rest of the network which is going to present the most problems.

The issue is 'de-skilling' and it's already happening, drivers of expresses can have quite long periods with nothing to do except keep watch so how do they maintain their attention focussed on spotting possible problems in the far distance, visibility permitting, and be 'up-to speed' in knowing what action to take? it was thought to be a factor in why Signallers failed to intervene when they could have done to prevent the Ladbroke Grove collision. If the routine work is automated how reasonable is it to expect a person sitting and watching to detect a rare serious problem?

I agree entirely, particularly from a signalling point of view (even spotting and dealing with potential regulating problems must be difficult when the computer's running the job). Surely that's a string argument for retaining Drivers in their current form though, and in full control of the train? The chap at the front of the express, whilst perhaps not doing very much between stations, is at least having to keep an eye on the speed of the train, check signals and watch for anything amiss - whereas if he's just there to watch the computer do it, it'll probably take a lot longer to respond to a potential hazard. ERTMS, or whatever the kit on the train that interfaces to it is called, can still supervise and intervene if necessary (just as TPWS and ATP do, with varying degrees of complexity, currently).

 

[notadriver]

A computer controlled train cannot be driven to break speed limits and take other unsafe actions in an attempt to keep to time.

British train drivers never speed nowadays either.

 

[island]

It's simple scientific fact that the basic driving task can not only be automated but done better by computer

I assume you can quote some scientific studies to back this up then?

 

[Robsignals]

But therein lies the issue. You find other posts to be hostile/insulting, but you don't see why the 'button pushing' post you made at the outset might be provocative or insulting to drivers?

We know there is way more to the job than button pushing and drivers have chipped in to point certain things out during this thread. The wrong routing issue for example happens way more than you seem to be aware of. The public are often oblivious because it is normally picked up by the driver and dealt with via a quick call to the signaller, it only gets noticed if the driver accepts the route. Of course it may not happen to a driver for 6 months or more, but that's where the trap lies. There are so many more things involved in being a competent driver than most people realise... but it's getting late on Christmas eve, so have a happy one.

Yes, 'button pushing' is an oversimplification - only a few power/brake levels now so might as well be buttons - to make a point. I'm making a reasoned argument but many replies aren't amounting to "can't be done, stupid to suggest it", by all means make reasoned points against otherwise best keep quiet.

Wrong routing is a human failure by the signaller now being eliminated by Automatic Route Setting and Regulation, anyway the train will have validated it's routing and confirmed with ARS that the correct route is set. All issues directly related to train operation are easily automated including constant monitoring of many more sensor inputs than a human could hope to do, the visual recognition systems on driverless cars are equally applicable to trains.

 

[Tomnick]

Wrong routing is a human failure by the signaller now being eliminated by Automatic Route Setting and Regulation, anyway the train will have validated it's routing and confirmed with ARS that the correct route is set.

Not always human error on the Signalman's part - especially with freight trains, some of which have numerous different schedules in the system, it's easy enough for it to be a problem with the schedule that's been called.

 

[Beveridges]

What would be the point in spending so much money automating trains when they would still need someone there in case things go wrong? Seems like a real waste of money to me. Also it would make the job extremely boring if you are doing nothing 90% plus of the time, only there in case something goes wrong? Don't see how computers would improve safety much when we already have TPWS which stops most collisions from happening. Also if it ever did come out it would be a very long time off seeing how much of a struggle it was to get ERTMS to work on a very quiet line in Wales.

Now one form of driving that I could never see go over to computers is Depot Driving. Too expensive to replace too few staff. Also a Computer may drive a train around a depot & avoid conflicting moves (and people), but it certainly won't be able to fuel a train, toilet tank a train, rectify faults, perform a full prep, unblock a toilet or pull a set of handpoints (unless more £millions is spent making them all power operated - pointless). A computer wouldnt be able to put a unit back together that has been split in half or into 3 pieces, a computer won't be able to put scotches under a unit then isolate its parking brakes. Completely pointless having a computer doing the depot drivers job when you'd still need someone onboard to do all the non driving manual work as the units move through the different parts of the depot. So Depot Driving will always be safe even in the unlikely event that mainline ever goes automatic.

 

[bronzeonion]

If the harsh punishments given by JR West were not there, the driver would not have felt the need to speed etc, that crash was definitely down managerial issues. And if you can't trust your workforce not to risk lives etc, well then there is definitely something wrong with humanity.

 

[Robsignals]

Quite a bit actually, who would drive a failed train, who would override failed safety systems? Fault find on the train and rectify it, who would do it in the middle it in the middle of nowhere? Who would deal with assistaning passengers etc the list goes on...pulled pass com and train comes to a half.

There's a reason why the top level ertms still retains a driver for mainline driving. There is so much more to being a train driver than just driving.

The control system would be duplicated or triplicated to be as reliable as a driver and eliminate ordinary control faults, other faults would probably require being rescued by another train which could also be done by the control system working with the central control which would also be able to talk to passengers.
--- old post above --- --- new post below ---

Surely that's a string argument for retaining Drivers in their current form though, and in full control of the train? The chap at the front of the express, whilst perhaps not doing very much between stations, is at least having to keep an eye on the speed of the train, check signals and watch for anything amiss - whereas if he's just there to watch the computer do it, it'll probably take a lot longer to respond to a potential hazard. ERTMS, or whatever the kit on the train that interfaces to it is called, can still supervise and intervene if necessary (just as TPWS and ATP do, with varying degrees of complexity, currently).

They are already monitored, other than for speed which can be set on 'cruise control', so drivers have little reason to pay much attention and human nature says they will start to lose interest.
--- old post above --- --- new post below ---

I assume you can quote some scientific studies to back this up then?

Which do you think can maintain constant unwavering attention, monitoring multiple sensor inputs and precisely calculating responses thousands of times a second, hour after hour? Care to cite any studies that support human over computer?

 

[barrykas]

They are already monitored, other than for speed which can be set on 'cruise control', so drivers have little reason to pay much attention and human nature says they will start to lose interest.

I challenge you to find "cruise control" on the vast majority of the current *MU fleets. Think you'll find it's mostly limited to locos, BICBW.

 

[Robsignals]

What would be the point in spending so much money automating trains when they would still need someone there in case things go wrong? Seems like a real waste of money to me. Also it would make the job extremely boring if you are doing nothing 90% plus of the time, only there in case something goes wrong? Don't see how computers would improve safety much when we already have TPWS which stops most collisions from happening. Also if it ever did come out it would be a very long time off seeing how much of a struggle it was to get ERTMS to work on a very quiet line in Wales.

They wouldn't need to be in the cab, they can be on passenger duties until there's a problem. There has already been at least one instance of a driver overiding TPWS because they didn't know they had SPADed and thought it was a faulty activation, can't happen with computer control.
--- old post above --- --- new post below ---

I challenge you to find "cruise control" on the vast majority of the current *MU fleets. Think you'll find it's mostly limited to locos, BICBW.

Then it's a proven system that could be universally fitted.

 

[455driver]

I think what's effectively been said here by drivers is that in fog instruments can "drive" the train without human visibility. Therefore the instruments could just drive the train. I think the biggest block to driverless trains isn't technology, its the unions.

What unions?
Are you saying the driverless trains will be in a union?

ASLEF wont exist because there wont be any train drivers, apparently!

 

[Robsignals]

Not always human error on the Signalman's part - especially with freight trains, some of which have numerous different schedules in the system, it's easy enough for it to be a problem with the schedule that's been called.

As the driver must know their destination now so the correct destination will be entered into the system which can select the best route taking account of live traffic conditions.
--- old post above --- --- new post below ---

What unions?
Are you saying the driverless trains will be in a union?

ASLEF wont exist because there wont be any train drivers, apparently!

As I've said there will be a member of staff on every train who's union will have the power to stop all trains.

 

[HSTEd]

What would be the point in spending so much money automating trains when they would still need someone there in case things go wrong?

So you think they pay on board catering/service staff the same amount as drivers?
They would only be driving very slowly in an emergency situation, they would also not require complicated route knowledge training for this reason.

Seems like a real waste of money to me. Also it would make the job extremely boring if you are doing nothing 90% plus of the time, only there in case something goes wrong?

They would not spend there time in the driving position if the train is under computer control and functioning normally, they would be serving customers or doing whatever it is the TOC wants them to.

Don't see how computers would improve safety much when we already have TPWS which stops most collisions from happening. Also if it ever did come out it would be a very long time off seeing how much of a struggle it was to get ERTMS to work on a very quiet line in Wales.

The ERTMS installation in Wales was done on the cheap and is ERTMS Level 2, which has all the costs of ERTMS and almost none of the benefits in maintenance terms. (It requires full track circuiting and the like which is the failure prone parts).

Now one form of driving that I could never see go over to computers is Depot Driving. Too expensive to replace too few staff. Also a Computer may drive a train around a depot & avoid conflicting moves (and people), but it certainly won't be able to fuel a train, toilet tank a train, rectify faults, perform a full prep, unblock a toilet or pull a set of handpoints (unless more £millions is spent making them all power operated - pointless). A computer wouldnt be able to put a unit back together that has been split in half or into 3 pieces, a computer won't be able to put scotches under a unit then isolate its parking brakes. Completely pointless having a computer doing the depot drivers job when you'd still need someone onboard to do all the non driving manual work as the units move through the different parts of the depot. So Depot Driving will always be safe even in the unlikely event that mainline ever goes automatic.

Why would you need a main line qualified driver to move a train around a depot?
The fitter could just plug a handheld unit into the MU jumpers on one end and move it around like that.
And depot driving is probably the easiest to automate because you need no predictive emergency cut-offs as they are moving so slowly in most cases that manual ones will suffice.

 

[NSEFAN]

Ultimately it is cost which determines if ATO is appropriate. It's the same reason that many locations are still using mechanical signalling. It is primitive by modern standards, but does the job sufficiently and the cost of upgrading wouldn't be recuperated quickly enough to justify the expense. If a line is upgraded, it then becomes cheaper overall to include the signalling in the infrastructure upgrade.

I have no doubt that computers can do driving better than people can. However, the cost of retro-fitting to the current network is currently not really worth it. The only places where ATO is a good option are on lines with extremely intense services, such as underground routes or the Thameslink/Crossrail cores. These are also enclosed environments, which further reduces the required complexity of the ATO. As has been mentioned, exposed mainlines have a lot of additional hazards compared to the tube.

As for staffing, there is a difference between driverless and unstaffed trains. I'm sure most passengers would prefer to see a human on board to help out in emergencies, even if most of the time it's the computer doing the hard work.

 

[notadriver]

HSTed - have you had many cab rides in a main line train ? You'd find it informative and interesting.

 

[Tomnick]

I think there's an increasing amount of fantasy in this thread - we all know that driverless trains are a possibility, but I think there's a minimal chance of seeing them introduced across a large part of the network in my lifetime (and I'd like to think I've got a few decades left yet!).

They would only be driving very slowly in an emergency situation, they would also not require complicated route knowledge training for this reason.

Drivers don't earn their money driving trains under normal conditions. They're paid such large amounts because of their rules knowledge to enable them to safely deal with just these out-of-course or emergency situations - when things will normally be moving slowly anyway (but a slow-moving train is just as capable of going somewhere it's not been authorised to go!) in the real world at present.

As the driver must know their destination now so the correct destination will be entered into the system which can select the best route taking account of live traffic conditions.

Possibly one day, I'm sure, but a long, long way off. You seem to be suggesting some means of instantly producing a VSTP path taking into account train length, necessary intermediate calls, traincrew relief, route restrictions and many more besides. Even then it's not unknown for Drivers to be unaware of a change to 'the plan' for their train, particularly when it's all going wrong.

 

[Metroland]

Whereas I think more automation is the way forward as a tool to help operation staff and cut costs, it's nowhere near as simple as it may seem.

Operating a real railway is multifaceted and still depends heavily on experienced operators applying rules and regulations, and well as using flexible iterated approaches to dealing with situations. The railway is as much about people as machines, it is full of legacy systems, and like any mode of transport is affected by freak occurrences (weather, failure, mishaps).

Automation needs to have a point. One point could be to save money, and this has been done with signalling systems over the last 100 years or more. Long sections of automatic semaphores were introduced as long ago as the 1920s.

Another point is to make systems more safe, and there is no question by and large machines are generally safer than humans at many (especially repetitive) tasks. They are not however infallible, because they are only as good as their designers at anticipating freak occurrences.

In the next 20 years I can see the railway using computers more for traffic regulation, data monitoring, systems performance. They could be brought in to manage engineering possession management more safely and effectively. For train driving itself, whereas we have got technology to do the nuts and bolts of driving trains, we are not quite there with doing all the duties of the driver (or signaller for that matter). It is much more than 'pressing buttons', and despite the 'simple as pressing a button' tag-lines of the 60s, which were somewhat economic with the truth, it will continue to be.

For the foreseeable future will still need someone on board the train, probably in the cab. The trains may well be semi-automatic in the future, but the step to full automation must tick a number of boxes on cost, safety, performance, public acceptance, simplification of working and ability to deal with freak and non-standard working arrangements, as well as failures, public interfacing etc.

There are wider questions over transport in general in future and what the railways are for. The steel wheel and rail has held its own with speed, capacity, comfort over the years and to an extent cost. Nevertheless, what other developments might there be in other forms of transport? This might have implications for the railway, but then with some many of these developments it needs to have public acceptance, be practically useful and serve a purpose in itself.

 

[Beveridges]

Why would you need a main line qualified driver to move a train around a depot?

Who said anything about a mainline qualified driver moving a train around a depot? Some TOC's have a specific grade to perform that task (plus the tanking, etc as the units move through the different stages around the depot)

The fitter could just plug a handheld unit into the MU jumpers on one end and move it around like that.
And depot driving is probably the easiest to automate because you need no predictive emergency cut-offs as they are moving so slowly in most cases that manual ones will suffice.

To automate depot driving - firstly ALL handheld points would have to go power operated plus with a Computer controlling all those as well as the trains. Secondly you'd need a different system to one that could be used for mainline, one dedicated to each individual depot. Thirdly you'd still need people to prep, fuel, toilet tank, fault find, unblock toilets, put split units back together, in fact so many manual tasks that automating Depot Driving really would be pointless and would hardly reduce the number of traincrew required on a Depot for an extremely high cost. You'd still need someone on the train to fuel/tank it when it gets to the fuel rig; and someone to shut all the windows before it moves through the wash plant; then someone to unblock the toilet when the train goes over the toilet apron; then someone to scotch it in the shed; or to prep it when it moves onto the sidings... I could go on forever with this. Yes they could pay lower wages for non-driving staff to do all the manual tasks if the Depot Driving goes automatic, but for the cost of bringing in a custom-made computer system designed just for the specific depot, and upgrading all the handpoints to power operated would not be worth it. Depot Driving will never go automatic the infrastructure on some depots is so poor it is not something which most companies wish to spend money on. You've got hand points that are extremely stiff and hard to pull, points that can split and derail trains, points that don't always fit tight when pulled, a fuel rig that spills out 1000's of litres of fuel and leaks diesel everywhere, I could go on forever with this list... the investment is not there on getting the very basic things right so what chance is there of spending many millions on state-of-the-art automatic systems in this sort of environment none whatsoever.

 

[maniacmartin]

An interesting note is that even the DLR is manual driven in the depot, as the cost to automate it was deemed too costly (but possible)

 

[Mintona]

Some serious jealousy coming across from certain posters in this thread, who I presume will not be happy until all drivers are out of a job.

 

[Latecomer]

Yes, 'button pushing' is an oversimplification - only a few power/brake levels now so might as well be buttons - to make a point. I'm making a reasoned argument but many replies aren't amounting to "can't be done, stupid to suggest it", by all means make reasoned points against otherwise best keep quiet.

You keep responding to my posts in this manner - I don't think anyone who has posted in this thread should be advised to "best keep quiet". Your original post was quite insulting in my view. You ignored the reasoned arguments you apparently seek for much of the time and you have displayed a lack of knowledge about what we currently do or may be called upon to do. Whilst a degree of further automation is inevitable most of the arguments you put forward talk about a utopian future which will not happen for a very long time (if ever) on the network as a whole. Whilst automation may improve, most of the automated services I have been on have an inferior ride quality compared to those of a skilled train driver (particularly when congestion occurs). In the meantime however your constant assertion that this or that can be done better by computer is demeaning to most of us who do the job and will be doing the job for many, many years to come. Even when a case is put forward (like the wrong route issue), you attempt to dismiss it because it can be elliminated in the future. We deal with now! We get paid for what we do now!

Because you don't appear to take these issues on board your declared 'admiration' from the outset appears glib when associated with the "press and go" or "button pushing" terminology you associate with our skills. In fact I would go further and say it's downright direspectful. There's an interesting debate to be had, but I suspect that some drivers might have already withrawn from this thread because whilst you made a provocative statement at the outset you seem to smart at any hint of someone taking umbrage with the line you are taking. Debate about the future is good (even though I disagree with your take on it), but to belittle what we do know with the terminology you have used because a few aspects of train handling might be improved by computer in years to come is bound to create an impassioned response. You surely knew that when you posted what you did?!

I think that I'll withdraw from this thread on that note because it's head versus brick wall time (although I won't ballot other drivers to do the same).

 

[Metroland]

http://www.futurerailway.org/RTS/About/Documents/RTS 2012 The Future Railway.pdf

This is the industry view of the future, it certainly envisages far more automation than is present, but not the complete elimination of key ops staff.

Advanced intelligent and automated traffic management systems
should progressively replace existing systems in control centres.
TSLG’s FuTRO project is developing the frameworks for the concepts,
requirements and architectures of next generation traffic management
systems. These systems should be dynamic and able to optimise the
use of the rail network, minimise delay, optimise traction energy use
and maintain train connections for passengers. Data from trains and the
infrastructure should predict where and when conflicts are likely to arise
and offer/implement solutions in real-time. Operating data should also
be used to feed automated long-term planning systems to optimise
train timetabling and infrastructure use. The FuTRO business case
estimates that the current value of benefits will be £200-£400m/year by
2035 derived from capacity increase, better information and operational
efficiencies.

ATO should be deployed widely across the network to provide
considerable benefits in standardising the driving operation. This is
particularly important as metro-style and urban/suburban operations
require predictable timekeeping to deliver intensive services. Combined
with intelligent traffic management systems, ATO could offer benefits
on other types of routes.

In-cab signalling is used instead of lineside signals and the only
traditional features of lineside signalling are point operating equipment
and level crossings. Signalling system designs are standardised and
the design, testing and commissioning procedures are automated.
The use of ATO is widespread across the network. Fully automatic
operation of trains is possible on some parts of the network

 

[bronzeonion]

The future is bleak then if any of that happens!

 

[DarloRich]

piece of p*ss tbh - you just sit on your rse looking out of the window pulling levers, pushing buttons and twiddling with knobs. Oh and drinking tea. Then going on strike, despite getting paid a bloody fortune. Easiest job in the world. Dont know they are born. Them and signal men.

Now you want to try working in an office. That is hard. In work for 9, away by 5, an hour for lunch, 5 days a week, having to answer the phone and send emails, going to meetings, reading and producing reports, sorting out all the operations staff mistakes, doing ALL of the work, getting none of the credit and getting paid less than lazy drivers AND trying to chat up the fit birds in the office. Now THAT is hard work!

Dont know they are born i tell you!

AND you dont have to be knocked up at 0230 am by a bloke wit tapper on the end ut pole to get down t'shed and get fire started under t'kettle!

THAT was is if you were lucky and had been promoted from cleaning dogs body t fireman! These days you just rock up at about 0800, grab a brew in your insulated mug, get on board, press a button and away you go!

Nowadays they let any old Tom, Dick or Henrietta drive trains. When i was a lad you Dad hat to be ont railways fut 50 year, drink with Bert the foreman downt railway club every night and be a decent inside forward fut works team before you even got a sniff oft footplate! Youse didn't even look at driver in thems days let alone think oft being one!

These days they make you take tests, understand rules andt regulations stuff, int old days we just used to make it up as we went along! Its not like you have t shovel coal fut 200 miles uphill 12 hours a day and keep pressure right int yon boiler is it.

I mean tests to drive a diesel or t'lectic!!!!!! How can it be? Engine does allt work! Int my day a driver was lucky to get £4 8sh & 6p a year and you weret grateful ft job! They get thousands now! ahm not jealous y understand just it aint like it uset be. Allt fun as gone out job now. They wont even let you gant bar before early turn!

Piece of p*ss these days ah tells yt!

 

[island]

Which do you think can maintain constant unwavering attention, monitoring multiple sensor inputs and precisely calculating responses thousands of times a second, hour after hour? Care to cite any studies that support human over computer?

You've made the assertion, it's up to you to prove you're right not for me to prove you're wrong.

 

[Robsignals]

Some serious jealousy coming across from certain posters in this thread, who I presume will not be happy until all drivers are out of a job.

Jealous of what? I worked in Signal Design, we control the trains with or without a driver. My original point was prompted by drivers saying they happily drive at full speed in zero visibility which seems to remove a major objection to driverless trains, it's spun off wildly since then even though I've agreed it won't happen for a long time...but, computing power is rising and cost falling at dramatic rates so what's impossible now may be possible sooner than we think.

For the third Boxing Day in a row there's a Tube strike in London but shops are reporting their best ever trade - no-one is indispensable, draw your own conclusions.

Latecomer - you seem intent on picking a fight, I'm not interested and have stopped reading your posts.

 

[cjp]

You've made the assertion, it's up to you to prove you're right not for me to prove you're wrong.

How many hours or days can you go without sleep?
Computers do not sleep.
With backup redundancy they perform and perform their program safely day after day after day.

We are fortunate that, thanks to the skills of the train operating staff and investigations into accidents, it will possible to program into a computer safe working parameters that will make safe computer operation possible eg DLR Victoria and Central lines - but not in my lifetime on Nationl Rail