Is route knowledge an outmoded concept?

[dk1]

Thanks for all that info.

My pleasure. Have a good evening.

 

[D1009]

German drivers do have route knowledge, but it is supported by the EBuLa concept - originally paper and now electronic - to which the driver refers whilst driving and which lists things like temporary speed restrictions. The radio extension to EBuLa supplies short term planned working timetables.

(Sent from my Mac not far from the Hamburg - Lübeck - Travemunde line...!)

EDIT: The debate about the advantages and disadvantages of speed or route based signalling will eventually fade away with the spread of ERTMS (if it is affordable) as this is essentially a speed-based system. The driver may be given indications of the route to be taken but ERTMS/ETCS essentially gives movement authority to a given point and therefore gives the driver indications of target speeds to be met at various points before the point where the movement authority expires.

Your quote from the undated German website refers specifically to special trains, unless something has been lost in translation. I'm trying to find out the level of route knowledge which is required of German drivers over routes they do not normally drive over, as this could be an issue when ETCS with cab signalling is widely implemented in the UK. I remember when the trial installation on the Cambrian started, there were issues with the way the required data is presented to drivers, and the potential for distraction from actual driving.

 

[coppercapped]

Your quote from the undated German website refers specifically to special trains, unless something has been lost in translation. I'm trying to find out the level of route knowledge which is required of German drivers over routes they do not normally drive over, as this could be an issue when ETCS with cab signalling is widely implemented in the UK. I remember when the trial installation on the Cambrian started, there were issues with the way the required data is presented to drivers, and the potential for distraction from actual driving.

I understand 'special trains' (from other sources) to be the equivalent of the UK's STP - Short Term Planned trains. These would mostly be freights but I suspect that the term also covers passenger trains replanned after a disruption. As I have been informed (admittedly some years ago!) drivers of passenger trains certainly need route knowledge, supported by EBuLa - but I am not sure about freight as there are still a large number of irregularly worked sidings in customers' premises and freight-only branches.

I would suggest that you have hit upon one of the major issues with ERTMS/ETCS. It is a politically inspired technical solution to an issue - loss of traffic and loss of market share in international traffic - which is organisational and managerial at root. It is not technical issues which have caused this loss but outmoded ways of working.

The traffic is there - anyone who has driven in Germany during the week knows that the Autobahnen are filled, wall to wall, with large artics from every country in Europe.

Squaring ERTMS/ETCS with the UK way of doing things without throwing the baby out with the bathwater is going to be very difficult. It's not as if it will enhance safety or traffic capacity to any significant extent.

 

[cjmillsnun]

I hope he/she spent it well.



So before GSMR how would they have stopped the train ?

NRN or CSR?

 

[coppercapped]

So before GSMR how would they have stopped the train ?

Waving their red knickers...?

 

[TUC]

So you're about 2 miles from the bridge.

Which way is the nearest "safe place" where you don't have to worry about further rail traffic?

Which way is the best way for potentially injured people to go to meet emergency services? Is there an obstruction or some sort of other hazard between the railway line and the nearest road? Is the route suitable for the barely-walking wounded? How do you know if the other tracks are active or not?

The answer to me would be "route knowledge"...

The answer is surely a rail version of Google Maps with that type of information included for the full length of the route.

 

[Hartington]

I'm one of those who believes technology can help. Can it help right now? Probably not. What would it take to allow it to help? Well first, NR need to get a handle on their estate. With out of date sectional appendices and the like any attempt at digitisation would be futile. They also need to determine where the GPS gaps are and find a reliable method of closing them (wiggly/leaky wire, balises or whatever).

Once that is in place a system to help drivers day to day and even to allow them to drive (probably on a cautionary basis) over routes they don't know then becomes possible.

Whether the cost of all that is worth while I don't know but one thing I am sure of is that NR needs to get a handle on their estate whatever that data is used for.

I can't help thinking the the unions should be thinking about this (are they?) with a view to understanding what the technology options are and then working out how this enhances the status of their members.

 

[Heartland]

Route knowledge has always been a key tool used by the Driver and Conductor. For the train operating company the cost of regular training and revision is an expense that they are keen to avoid and removal of this tuition would increase profits. Yet as the digital railway is now on the horizon, there may be available with on board computers all the necessary revisions for the driver and conductor to use.

 

[SpacePhoenix]

Apart from braking points, is the route knowledge for both drivers and guards identical?

 

[Johncleesefan]

Apart from braking points, is the route knowledge for both drivers and guards identical?

I would have thought opposite, guards learn the route from the back cab, so essentially in reverse. They don't need to learn signals necessarily either, I believe they need speeds though

 

[najaB]

Apart from braking points, is the route knowledge for both drivers and guards identical?

Guards also need to know the correct dispatch points at stations.

 

[Johncleesefan]

Guards also need to know the correct dispatch points at stations.

I should edit nd say guards do need to know relevant signals ie platform and starter sigs

 

[LAX54]

I take it none of you ever fly. And what do you do when you have to drive your car sometimes you've never been before?

Yes, I know, neither is the same as driving a train but let's concentrate on flying for a moment. How do they find their way around an airport? A map and the mark one eyeball! There are the equivalent of traffic lights and filters to tell them which way to go (sometimes). If the route changes in the air (yes it happens) they reprogram the autopilot and, you know what, they make sure they know where they're going. You know what? When they happen to misprgram they just land, safely, at another airport.

There are a few airports where you have to go with an experienced pilot your first time (Gibraltar, Madeira and San Francisco to name three) but generally speaking if you haven't been there before you just follow the chart.

You don't think twice about getting in your car to drive somewhere you've never been before and you could come across a slippery road and you deal with it. You get on a plane. Why is a train so different?

Can't recall I saw a train do a 3 point turn as it took the wrong route ! come to that, stop in a lay by to look at a map !

 

[dk1]

Can't recall I saw a train do a 3 point turn as it took the wrong route ! come to that, stop in a lay by to look at a map !

Exactly. Let's not forget that at speed a train cannot simply stop either. Some very scary & Ill thought out replies on here.

 

[fowler9]

Exactly. Let's not forget that at speed a train cannot simply stop either. Some very scary & Ill thought out replies on here.

Yeah, I'm not sure driving a car or flying a plane are remotely comparable. All have their unique challenges and dangers.

 

[XDM]

Exactly. Let's not forget that at speed a train cannot simply stop either. Some very scary & Ill thought out replies on here.

Let's risk being branded "ill thought out", or worse epithets.
What is the risk of ECML drivers who do not sign the diversionary route via Hertford taking that route when the main line between Alexandra Palace & Stevenage is blocked?. In the dark, far too risky if they drive at what used to be called line speed. In daylight still too risky. If they miss an advance PSR/TSR warning,or see a signal late it would be catastrophic. But if they drove at 45 mph max in daylight & 30 mph in darkness, TPWS & the low speeds would eliminate the risks, or would they?
I & some driver colleagues,but not all, think it would be perfectly safe. And most of those I mentioned it to would do it if they were guaranteed no "please explain" or enquiry, if they Spadded. Plus the right to drive at below 45 mph if they were uneasy. What risks are there at these low speeds with TPWS? I do not think there are any. The benefits are obvious, that more main line long distance services could be run if the line was blocked at say Hitchin. The same principle could apply everywhere on the railway with higher or lower top speed restrictions. I can see it being vigorously opposed,but the opponents motives should be challenged.

 

[Hartington]

A train can't stop quickly? Neither can a plane; in fact it's probably worse.

On the night the first Gulf War started a pilot I know was operating Heathrow to Jeddah. As he approached the Egyptian coast he was told that Saudi airspace was closed. He decided to turn around and go back to Athens (on the basis that there were crews there who could take over. He reckons it took about 20-30 mins to get turned around (no, not a three point turn). It's not just about doing a 180 degree turn you have to change altitude, probably at least twice. Having done that he looked at his fuel situation and realised he had enough to get back to Heathrow. So he did that and parked on the stand next to the one he departed from. "Never took so long to go such a long distance" he said.

I fear many of you take things much to literally. Can a GPS you use in the car be of any use in a driving a train? Of course not; anymore than a car GPS can be used safely in a lorry (I wonder how many bridge bashes are the result of a lorry using a car GPS?). Can a GPS, with other enhancements, provide drivers with support to improve their safety and, under defined circumstances, drive routes they don't sign? Yes. It doesn't exist yet, some of you will have retired by the time NR get their act together, digitise their asset base, install the extra systems trackside that will be required and generally understand all about braking points, different braking points in different conditions, dispatch points and all the other things you currently know that "systems" don't.

This isn't about the ATO, this is about enhancing your knowledge to do your job more safely, providing you with support to do your job more safely. But it will also allow you to provide a better service to your customers in times of disruption. I know some of you drive old trains and you think the "GPS system" for want of a better name, can't be made to work with the systems that already exist on your trains. It doesn't have to. It's data to help you. It can sit there, in the cab (it will need a mount and power), and tell you where the next signal is; it won't tell you the aspect, simply where the signal is and where the next junction is and what the next speed limit (including TSRs) is and where it is and the current gradient and the next gradient after that etc etc. I know you know all that but this isn't about replacing you, a properly designed system will enhance your skills, not replace them.

Now, of course, the issue is how long will it take NR to understand what it needs to do. I strongly suspect that they simply don't know and, in the end, something like ATO will actually overtake the kind of systems we're talking about here.

I also think the systems are feasible but I don't believe the industry as a whole has the nouse to actually implement anything.

 

[fowler9]

A train can't stop quickly? Neither can a plane; in fact it's probably worse.

On the night the first Gulf War started a pilot I know was operating Heathrow to Jeddah. As he approached the Egyptian coast he was told that Saudi airspace was closed. He decided to turn around and go back to Athens (on the basis that there were crews there who could take over. He reckons it took about 20-30 mins to get turned around (no, not a three point turn). It's not just about doing a 180 degree turn you have to change altitude, probably at least twice. Having done that he looked at his fuel situation and realised he had enough to get back to Heathrow. So he did that and parked on the stand next to the one he departed from. "Never took so long to go such a long distance" he said.

I fear many of you take things much to literally. Can a GPS you use in the car be of any use in a driving a train? Of course not; anymore than a car GPS can be used safely in a lorry (I wonder how many bridge bashes are the result of a lorry using a car GPS?). Can a GPS, with other enhancements, provide drivers with support to improve their safety and, under defined circumstances, drive routes they don't sign? Yes. It doesn't exist yet, some of you will have retired by the time NR get their act together, digitise their asset base, install the extra systems trackside that will be required and generally understand all about braking points, different braking points in different conditions, dispatch points and all the other things you currently know that "systems" don't.

This isn't about the ATO, this is about enhancing your knowledge to do your job more safely, providing you with support to do your job more safely. But it will also allow you to provide a better service to your customers in times of disruption. I know some of you drive old trains and you think the "GPS system" for want of a better name, can't be made to work with the systems that already exist on your trains. It doesn't have to. It's data to help you. It can sit there, in the cab (it will need a mount and power), and tell you where the next signal is; it won't tell you the aspect, simply where the signal is and where the next junction is and what the next speed limit (including TSRs) is and where it is and the current gradient and the next gradient after that etc etc. I know you know all that but this isn't about replacing you, a properly designed system will enhance your skills, not replace them.

Now, of course, the issue is how long will it take NR to understand what it needs to do. I strongly suspect that they simply don't know and, in the end, something like ATO will actually overtake the kind of systems we're talking about here.

I also think the systems are feasible but I don't believe the industry as a whole has the nouse to actually implement anything.

I still don't agree. the whole point of the thread was along the lines of questioning if route knowledge is obsolete. I don't think it is or ever will be, there are a number of plane crashes which have involved pilots relying too much on automated systems and either trusting faulty automated systems or not knowing what to do when the automated systems stop working. I don't think the thread was asking if technology could possibly assist which you seem to be leaning towards.

 

[LAX54]

A train can't stop quickly? Neither can a plane; in fact it's probably worse.

On the night the first Gulf War started a pilot I know was operating Heathrow to Jeddah. As he approached the Egyptian coast he was told that Saudi airspace was closed. He decided to turn around and go back to Athens (on the basis that there were crews there who could take over. He reckons it took about 20-30 mins to get turned around (no, not a three point turn). It's not just about doing a 180 degree turn you have to change altitude, probably at least twice. Having done that he looked at his fuel situation and realised he had enough to get back to Heathrow. So he did that and parked on the stand next to the one he departed from. "Never took so long to go such a long distance" he said.

I fear many of you take things much to literally. Can a GPS you use in the car be of any use in a driving a train? Of course not; anymore than a car GPS can be used safely in a lorry (I wonder how many bridge bashes are the result of a lorry using a car GPS?). Can a GPS, with other enhancements, provide drivers with support to improve their safety and, under defined circumstances, drive routes they don't sign? Yes. It doesn't exist yet, some of you will have retired by the time NR get their act together, digitise their asset base, install the extra systems trackside that will be required and generally understand all about braking points, different braking points in different conditions, dispatch points and all the other things you currently know that "systems" don't.

This isn't about the ATO, this is about enhancing your knowledge to do your job more safely, providing you with support to do your job more safely. But it will also allow you to provide a better service to your customers in times of disruption. I know some of you drive old trains and you think the "GPS system" for want of a better name, can't be made to work with the systems that already exist on your trains. It doesn't have to. It's data to help you. It can sit there, in the cab (it will need a mount and power), and tell you where the next signal is; it won't tell you the aspect, simply where the signal is and where the next junction is and what the next speed limit (including TSRs) is and where it is and the current gradient and the next gradient after that etc etc. I know you know all that but this isn't about replacing you, a properly designed system will enhance your skills, not replace them.

Now, of course, the issue is how long will it take NR to understand what it needs to do. I strongly suspect that they simply don't know and, in the end, something like ATO will actually overtake the kind of systems we're talking about here.

I also think the systems are feasible but I don't believe the industry as a whole has the nouse to actually implement anything.

and none of that is any good if you have not got a clue where you are supposed to be going, its not good just putting a route into a rail version of GPS, route knowledge also applies to what train you are driving, a class 4 may run via route A one day, but due to loadings may have to run via route b, or c or d or more,
With GPS being followed, if you don't know the route, you will end up in the wrong place quite easily, nothing can ever replace human route knowledge
Its nothing to do with NR getting their act together.
Yes Commercial Pilots may well go to an alternate airport they don't know, guided by ATC on the ground giving them details.
There can be, and is too much 'automation' reliance on computers, internet connections, networks etc, far too much, to the extent that when it fails, whatever is going on related to that just stops.

 

[philthetube]

A train can't stop quickly? Neither can a plane; in fact it's probably worse.

On the night the first Gulf War started a pilot I know was operating Heathrow to Jeddah. As he approached the Egyptian coast he was told that Saudi airspace was closed. He decided to turn around and go back to Athens (on the basis that there were crews there who could take over. He reckons it took about 20-30 mins to get turned around (no, not a three point turn). It's not just about doing a 180 degree turn you have to change altitude, probably at least twice. Having done that he looked at his fuel situation and realised he had enough to get back to Heathrow. So he did that and parked on the stand next to the one he departed from. "Never took so long to go such a long distance" he said.

I fear many of you take things much to literally. Can a GPS you use in the car be of any use in a driving a train? Of course not; anymore than a car GPS can be used safely in a lorry (I wonder how many bridge bashes are the result of a lorry using a car GPS?). Can a GPS, with other enhancements, provide drivers with support to improve their safety and, under defined circumstances, drive routes they don't sign? Yes. It doesn't exist yet, some of you will have retired by the time NR get their act together, digitise their asset base, install the extra systems trackside that will be required and generally understand all about braking points, different braking points in different conditions, dispatch points and all the other things you currently know that "systems" don't.

This isn't about the ATO, this is about enhancing your knowledge to do your job more safely, providing you with support to do your job more safely. But it will also allow you to provide a better service to your customers in times of disruption. I know some of you drive old trains and you think the "GPS system" for want of a better name, can't be made to work with the systems that already exist on your trains. It doesn't have to. It's data to help you. It can sit there, in the cab (it will need a mount and power), and tell you where the next signal is; it won't tell you the aspect, simply where the signal is and where the next junction is and what the next speed limit (including TSRs) is and where it is and the current gradient and the next gradient after that etc etc. I know you know all that but this isn't about replacing you, a properly designed system will enhance your skills, not replace them.

Now, of course, the issue is how long will it take NR to understand what it needs to do. I strongly suspect that they simply don't know and, in the end, something like ATO will actually overtake the kind of systems we're talking about here.

I also think the systems are feasible but I don't believe the industry as a whole has the nouse to actually implement anything.

A train can't steer out of trouble

 

[Dryce]

A train can't steer out of trouble

From an engineering perspective that means less to worry about.

It makes automation easier because it limits the range of control decisions that can be taken by on board systems.

 

[Dryce]

and none of that is any good if you have not got a clue where you are supposed to be going, its not good just putting a route into a rail version of GPS, route knowledge also applies to what train you are driving, a class 4 may run via route A one day, but due to loadings may have to run via route b, or c or d or more,

You don't need GPS on a rail network.

(You don't need GPS on an aircraft either for that matter.)

You can tell where you are on the majority of the rail system with an accelerometer and a clock. And for the bits where you can't figure that then you can use some trackside or on track markers that the equipment can understand.

Add a database to that and you have route knowledge in a black box.

Whether you choose to use that to assist or replace a driver is another question.

But the tech is not hard. And how reliable GPS is or is not is a distraction.

In aviation the automation is criticised as somehow undermining the pilots. The reality is that without it we'd probably think air crashes were not news worthy - we'd have aircraft hitting the ground a lot more frequently. The situations where things go wrong are edge cases - they don't undermine the fundamentals - and in a proportion of cases it works the other way - the human in the loop is the problem.

 

[jj1314]

You don't need GPS on a rail network.

...

You can tell where you are on the majority of the rail system with an accelerometer and a clock. And for the bits where you can't figure that then you can use some trackside or on track markers that the equipment can understand.

That's all well and good, but what about drift? Dead reckoning is only so useful. Even using a commercial IMU, I've had scenarios where my position estimate has become unusable within a few seconds when working on flat calm seas near cliffs. I'd imagine a swaying train has similar limitations.

I'd imagine a high number of trackside beacons would be required as a bare minimum.

 

[najaB]

That's all well and good, but what about drift? Dead reckoning is only so useful.

This idea aside, this is a problem that will need to be solved if ETCS Level 3 is ever to be a reality. I suspect the solution will be a combination of GPS, dead reckoning (e.g. counting wheel revolutions) and fixed beacons.

 

[dviner]

This idea aside, this is a problem that will need to be solved if ETCS Level 3 is ever to be a reality. I suspect the solution will be a combination of GPS, dead reckoning (e.g. counting wheel revolutions) and fixed beacons.

Most of the "idiots guide to ETCS" that I've seen make no mention of GPS, but tend to refer to odometry and radar/lidar plus beacons/balises.

How accurate this is to real life has to be seen.

As an intellectual exercise - why add a dependency on GPS, with all the complications added by tunnels, deep cuttings, and sub-surface stations when a combination of counting wheel revs and fixed beacons would work on the whole network? Inaccuracies created by trips to the wheel lathes and slides would be counteracted by the fixed balises saying "you are here".

 

[theironroad]

Whenthe computer thinks it know better than people...

http://www.canberratimes.com.au/goo...-leaves-pilots-powerless-20170510-gw26ae.html

Thanks for the link. Interesting and very sobering article.

 

[ComUtoR]

As an intellectual exercise - why add a dependency on GPS, with all the complications added by tunnels, deep cuttings, and sub-surface stations when a combination of counting wheel revs and fixed beacons would work on the whole network? Inaccuracies created by trips to the wheel lathes and slides would be counteracted by the fixed balises saying "you are here".

As an intellectual exercise - Why are we theorising something that already exists and is currently in passenger service ?

 

[najaB]

As an intellectual exercise - why add a dependency on GPS, with all the complications added by tunnels, deep cuttings, and sub-surface stations when a combination of counting wheel revs and fixed beacons would work on the whole network? Inaccuracies created by trips to the wheel lathes and slides would be counteracted by the fixed balises saying "you are here".

If it's available basically for free, it might be useful as a cross-check.

 

[dviner]

As an intellectual exercise - Why are we theorising something that already exists and is currently in passenger service ?

Being a layman, I wasn't aware that it is was - hence "intellectual exercise". So, that part's done, just need to get the transfer of movement authorities by GSM-R sorted out...

If it's available basically for free, it might be useful as a cross-check.

Something else to go wrong, need to interface it in, which is the definitive position data source? Does it leave the depot if the GPS is down?

 

[Dryce]

I'd imagine a high number of trackside beacons would be required as a bare minimum.

The thing with the railway is that it's fixed infrastructure - that gives you a head start.

There are a multitude of ways to get fixes. As long as you have a map. Lateral acceleration can be used to fix against the map as long as the track changes direction frequently enought. These days a camera system could be used to read existing visual signs and signals to confirm fixes - or use specialised markers (railway QR code markers) - or use something non-optical like RFID. You can count the rotations on undriven wheels. Or customised transmitters. There's GPS as a further backup. Cellular radio signal triangulation.