Level crossing stupidity...

[Dagworth]

Eeep!  I wouldn't want anything safety critical like barrier gates to be visible on "The Internet of Things"!

 

Imagine some plonker getting into the system and raising the barriers as a train approaches?  The railway system is reduced to allowing software that is possibly inadequately/incorrectly specified and imperfectly implemented to control vital infrastructure without any guarantee that it can't be penetrated by malicious idiots.

 

Its best to have a chap in a box, with a great big capstan wheel to control the gates.  Then we'd know who to blame when things go wrong...

as opposed to some plonker pressing the button to raise the barrier, and then driving his tractor into the path of the train?

 

Andi

[martin_wynne]

The object of POGO seems to be to avoid the user having to cross to open and close the opposite gate before/after use.

 

How about simply slaving the gates together? Make the user open the first one manually, and the opposite one follows suit under power? Likewise when closing the one behind you, the opposite one closes too?

 

No push buttons needed, no new signs or operating procedures, simply a self-evident powered aid to gate users, and nothing more. And probably less expensive -- add a gate position sensor and servo drive, and a trigger on the gate handle to release the clutch on the gate drive. Remove need for new signs, push buttons, any change to existing systems.

 

Martin.

Edited November 13, 2017 by martin_wynne

[96701]

The object of POGO seems to be to avoid the user having to cross to open and close the opposite gate before/after use.

 

How about simply slaving the gates together? Make the user open the first one manually, and the opposite one follows suit under power? Likewise when closing the one behind you, the opposite one closes too?

 

No push buttons needed, no new signs or operating procedures, simply a self-evident powered aid to gate users, and nothing more. And probably less expensive -- add a gate position sensor and servo drive, and a trigger on the gate handle to release the clutch on the gate drive. Remove need for new signs, push buttons, any change to existing systems.

 

Martin.

I'm struggling to see how that would work. You'd need 2 powered gates that could both be manually operated and slaved together. No mean technological feat.

[martin_wynne]

I'm struggling to see how that would work. You'd need 2 powered gates that could both be manually operated and slaved together. No mean technological feat.

 

Not really. A trigger on the gate handle releases that gate from its drive and engages the drive on the opposite one. Then a position-sensor on the gate as you open or close it replicates the position on the other gate.

 

Or of course in steam days, a system of rods and cranks links the two gates, and the whole thing relies on the user to work two gates at once. Then you also save the cost of providing a power supply, weatherproof cabinets, and all the rest of it. If properly maintained with good gate bearings and seals there is no reason that wouldn't work. Anyone capable of opening a typical field gate can exert a significant moment on the end of an 8ft lever, sufficient to operate two properly engineered gates very easily.   

 

Martin.

[Coryton]

Instructions on use are everywhere, if people fail to read them, then it is their problem.

 

 

While trains (and their occupants) usually come out best in an argument with a vehicle on a level crossing, it isn't always the case.

 

When I'm on a train I'd rather any crossings it passes over take into account how people actually behave not how we'd like them to.

[96701]

Not really. A trigger on the gate handle releases that gate from its drive and engages the drive on the opposite one. Then a position-sensor on the gate as you open or close it replicates the position on the other gate.

 

Or of course in steam days, a system of rods and cranks links the two gates, and the whole thing relies on the user to work two gates at once. Then you also save the cost of providing a power supply, weatherproof cabinets, and all the rest of it. If properly maintained with good gate bearings and seals there is no reason that wouldn't work. Anyone capable of opening a typical field gate can exert a significant moment on the end of an 8ft lever, sufficient to operate two properly engineered gates very easily.   

 

Martin.

I'm still at a bit of a loss. How do you propose a trigger on one gate disengages the drive on that gate and then engages the drive on the other gate? Properly engineered still requires proper maintenance. Installation of novel equipment requires approval which requires proper design, or are you really suggesting gates that close across the railway as the rods, cranks and wheels used to?

[martin_wynne]

I'm still at a bit of a loss. How do you propose a trigger on one gate disengages the drive on that gate and then engages the drive on the other gate?

 

A trigger switch. Some cable. A couple of relays, Solenoid-operated clutches on the gate gearboxes. Servomotors. You want me to provide the actual part numbers here and now?

 

Martin.

[roythebus]

All very well but not all crossings are near any power source, and "the railway" wouldn't allow solar panels to work safety critical equipment would they?

[JDW]

At first, I thought that was quite a good idea Martin, but the more I think about it, I'm not sure.  I like the theory, and can see how it could be done electrically - although an old fashioned idea of pulleys etc would probably be much more reliable in operation.  On reflection, I think it might be an idea that's come after it's time, and solutions involving interlocking, signalbox involvement by phone etc, are more likely to happen.  Whether that's right or wrong, good or bad, I'll not pass judgement.  I do think your idea has merit though, especially in remote areas where there is no power supply, a simple system which mechanically opens and closes both gates (either onto or away from the track) would be achievable and remove the need to cross and re-cross, which in turn speeds up the crossing, reduces time spent by the user on the tracks, and if surely if it were simply an amended way of opening and closing the gates would need much less in the way of new rules and regulations and procedures than would some complex system that is interlocked with signalling or needs electronics etc.

[96701]

A trigger switch. Some cable. A couple of relays, Solenoid-operated clutches on the gate gearboxes. Servomotors. You want me to provide the actual part numbers here and now?

 

Martin.

Not really Martin. My point is that your idea has probably been tried before, but by the time that you have interconnected two gates across a heavy rail line and installed clutches, rods, levers and whatever, the costs are rising all the time, generally beyond the point of cost effectiveness for a user worked crossing.

[Geoff Endacott]

If a raised barrier can give the false impression that the line must be clear, wouldn't it be safer to not have any barriers?

 

Geoff Endacott

[martin_wynne]

You want me to provide the actual part numbers here and now?

 

Here's a small electromagnetic clutch which might be sufficient for a gate-drive motor. 8mm shaft.

 

Here's a manufacturer of industrial clutches and couplings: http://www.djautomation.net/basic.html

 

For a cheap and cheerful solution, a centrifugal clutch and chain drive might work.

 

There could be a safety issue with this idea. If two people attempt to open opposite gates simultaneously, and one releases their trigger, the gate they are holding becomes powered. This needs some further thought.

 

That problem doesn't arise with the all-mechanical, no power, no batteries, works in the rain, approach. Unfortunately such solutions are regarded as old-hat nowadays, whereas with modern engineering they could save the planet. Just googling now for waterwheel bearings.

 

Martin. 

 

Martin.

[martin_wynne]

If a raised barrier can give the false impression that the line must be clear, wouldn't it be safer to not have any barriers?

 

In that case, you get down on your knees and pray:

 

 https://goo.gl/maps/58HV8GxnQtF2

 

Martin.

[Geoff Endacott]

In that case, you get down on your knees and pray:

 

...or have a red light and a green light.

 

Geoff Endacott

[RJS1977]

Here's a small electromagnetic clutch which might be sufficient for a gate-drive motor. 8mm shaft.

 

Here's a manufacturer of industrial clutches and couplings: http://www.djautomation.net/basic.html

 

For a cheap and cheerful solution, a centrifugal clutch and chain drive might work.

 

There could be a safety issue with this idea. If two people attempt to open opposite gates simultaneously, and one releases their trigger, the gate they are holding becomes powered. This needs some further thought.

 

That problem doesn't arise with the all-mechanical, no power, no batteries, works in the rain, approach. Unfortunately such solutions are regarded as old-hat nowadays, whereas with modern engineering they could save the planet. Just googling now for waterwheel bearings.

 

Martin. 

 

Martin.

 

No matter how the gates are linked to each other, it still doesn't solve the problem of people opening them when they shouldn't.

[martin_wynne]

No matter how the gates are linked to each other, it still doesn't solve the problem of people opening them when they shouldn't.

 

That applies no matter how the gates are arranged, linked, not linked, replaced with a length of rope, or a broken milk crate, or a man with a red flag. If it is a "user worked" crossing, it is by definition relying on the user to work it.

 

Martin.

[martin_wynne]

A possible solution would be to add an additional track as a siding, and park a train of rusty coal wagons on it. And only agree to move them if they ring up and ask very nicely to use the crossing.

 

Martin.

[martin_wynne]

A modification to the existing POGO equipment would be to add a memory chip to the push button unit, containing a list of say 5,000 random code numbers. The signaller has a copy of the list.

 

You ring the signaller, when/if it's safe he gives you the next code number on the list. You enter it on a number pad and the gates open. Any other code and they don't. Each code can only be used once, until it reaches the end of the list when it starts again from the beginning. A 4-figure code would be enough -- that gives you 10,000 random codes from 0000-9999 to try and guess. Easier to ring for the correct one.

 

If the code doesn't work, it means someone asked for a code but never used it. In which case you ring back, and the signaller gives the previous code on the list again.

 

Hopefully the number pad would be close to the phone, so the code can be entered as the signaller speaks to you, for those who can't remember a 4-figure number for more than 5 seconds. Also the signaller could hear the beeps as the code is entered, and thereby know the code is being used immediately, and that the user is not getting a code in advance, to come back and use it half an hour later, or give it to someone else on their mobile.

 

It would be defeated by someone leaving the gates open, but that applies to the system now, and is a serious offence already.

 

Martin.

Edited November 14, 2017 by martin_wynne

[Hroth]

...or have a red light and a green light.

 

Geoff Endacott

I'm just wondering how well that would work considering what the more selfish road user does when a set of traffic lights changes to red in front of them.  Although I am amazed at how well roadworks traffic lights are observed.  Perhaps its due to all those workmen/witnesses standing about?

[FPH 603]

An example of level crossing stupidity:

 

1992 TV Series 'Police Rescue' Episode 2 of Season 2 - Off the Track: https://m.youtube.com/watch?v=WAeHsm8NOpw

 

Watch from around 1:07 to 2:10 for the whole scene, but if you want to see the fool get hit go forward to about 1:45. Once you see this, put on this music: https://m.youtube.com/watch?v=ZnHmskwqCCQ

Edited November 14, 2017 by DoubleDeckInterurban

[stewartingram]

I'm just wondering how well that would work considering what the more selfish road user does when a set of traffic lights changes to red in front of them.  Although I am amazed at how well roadworks traffic lights are observed.  Perhaps its due to all those workmen/witnesses standing about?

I'm not so sure about that. Every time I come across roadworks lights people just keep going through when the light goes to red.

 

Stewart

Edited November 14, 2017 by stewartingram

[The Stationmaster]

A modification to the existing POGO equipment would be to add a memory chip to the push button unit, containing a list of say 5,000 random code numbers. The signaller has a copy of the list.

 

You ring the signaller, when/if it's safe he gives you the next code number on the list. You enter it on a number pad and the gates open. Any other code and they don't. Each code can only be used once, until it reaches the end of the list when it starts again from the beginning. A 4-figure code would be enough -- that gives you 10,000 random codes from 0000-9999 to try and guess. Easier to ring for the correct one.

 

If the code doesn't work, it means someone asked for a code but never used it. In which case you ring back, and the signaller gives the previous code on the list again.

 

Hopefully the number pad would be close to the phone, so the code can be entered as the signaller speaks to you, for those who can't remember a 4-figure number for more than 5 seconds. Also the signaller could hear the beeps as the code is entered, and thereby know the code is being used immediately, and that the user is not getting a code in advance, to come back and use it half an hour later, or give it to someone else on their mobile.

 

It would be defeated by someone leaving the gates open, but that applies to the system now, and is a serious offence already.

 

Martin.

 

I don't see how that creates any extra safety apart from the possibility of a crossing user opening the gates when there is a train coming because all it is doing is imposing any extra workload on the Signalman(ler) who could well have have a dozen such crossings in a section where he has no idea where a train actually happens to have got to although he knows it is in that section.  So all you are doing is imposing an extra step which might to some users imply greater security/safety for their use of the crossing than is actually the case.

 

The real answer is that we should get back to talk about occupation and accommodation crossings recognising what they are instead of keeping on thinking in terms that regard them as 'level crossings'. The only similarity they have to a level crossing is that they cross the railway line on the level - at that point similarity ceases because they are basically outwith the railway's control and rely to a very large extent entirely on user discipline rather than railway operational discipline and railway safety systems.  In many cases it isn't even possible for a Signalman, if a signalbox can be contacted, to actually say if it is safe to cross - all that he/she can say is 'wait until such & such a train has passed' and that could involve a wait of 10 minutes or more; and a wait that long is more than enough to tempt the crossing user to take a chance or assume the train has already passed having waited for what seems an age.

 

The alternative from a railway viewpoint is extra money on track detection and signalbox display systems so that a more positive or immediate answer might be given and that in turn increase workload which might mean an increased ongoing cost as well as the obvious extra maintenance cost.  The alternative from the user's viewpoint is simply to get rid of the crossing, which also affords some benefit to the railway of course, but we are in a very different financial ball game from closing a level crossing (in the correct meaning of that term.  For example we have an occupation crossing on our local branch, about a couple of miles from my front door as it happens so at times I hear trains blowing for the crossing approach. Notwithstanding relatively good sighting (which could be improved by felling a fair amount of lineside jungle) and low linespeed (50mph) the crossing is listed as 'High Risk' by NR.  On one side of the railway the landowners property extends right up to a very lightly used lane which crosses the railway by an overbridge only a few hundred yards from the crossing.  On the other side of the line he could easily afford to buy the small piece of land between his property and the bridle road - so at the cost of a couple of gates, a small land purchase and some minor earthworks the crossing could be eliminated - the cost would no doubt be considerably less than the amount of money the landowner spent last year on creating his second polo pitch and it would certainly be massively less than the value of any one of the polo ponies he takes across the crossing.  Totally different circumstances of use and closure possibility than a public level crossing and more typical (apart from the polo ponies) of the situation with many occupation etc crossings.

 

When HSTs were first introduced several dozen such crossing on the GWML were closed, literally, overnight following a careful study of their use and ready alternatives and none of the closures involved a need for bridgeworks.  Separate problem from level crossings and often much simpler solutions are readily available if people care to look.

[Coryton]

. Notwithstanding relatively good sighting (which could be improved by felling a fair amount of lineside jungle) and low linespeed (50mph) the crossing is listed as 'High Risk' by NR.  

 

I'm not sure how these risk figures are calculated.

 

There is a foot level crossing (not occupation) near me which is categorised as "collective risk" 2/13 (where 1 is the highest).

 

It has extremely good visibility in both directions and the line speed is, I think, 45 mph but it is very rare that a train reaches that because it is between two closely spaced stations (and also close to one end of the line). Even if the train isn't stopping you get a lot of warning that it's coming.

 

Perhaps 2 is as low as it gets without some kind of indication of when a train is coming, but I have seen far scarier foot level crossings (including some I really wouldn't want to use) and can't think of many that I've used that were safer.

 

It's also apparently used by passenger and freight trains. Perhaps that includes the autumn RHHTs because that's the only non passenger service that makes its way down there (apart from perhaps a track recording train)

[96701]

I don't see how that creates any extra safety apart from the possibility of a crossing user opening the gates when there is a train coming because all it is doing is imposing any extra workload on the Signalman(ler) who could well have have a dozen such crossings in a section where he has no idea where a train actually happens to have got to although he knows it is in that section.  So all you are doing is imposing an extra step which might to some users imply greater security/safety for their use of the crossing than is actually the case.

 

The real answer is that we should get back to talk about occupation and accommodation crossings recognising what they are instead of keeping on thinking in terms that regard them as 'level crossings'. The only similarity they have to a level crossing is that they cross the railway line on the level - at that point similarity ceases because they are basically outwith the railway's control and rely to a very large extent entirely on user discipline rather than railway operational discipline and railway safety systems.  In many cases it isn't even possible for a Signalman, if a signalbox can be contacted, to actually say if it is safe to cross - all that he/she can say is 'wait until such & such a train has passed' and that could involve a wait of 10 minutes or more; and a wait that long is more than enough to tempt the crossing user to take a chance or assume the train has already passed having waited for what seems an age.

 

The alternative from a railway viewpoint is extra money on track detection and signalbox display systems so that a more positive or immediate answer might be given and that in turn increase workload which might mean an increased ongoing cost as well as the obvious extra maintenance cost.  The alternative from the user's viewpoint is simply to get rid of the crossing, which also affords some benefit to the railway of course, but we are in a very different financial ball game from closing a level crossing (in the correct meaning of that term.  For example we have an occupation crossing on our local branch, about a couple of miles from my front door as it happens so at times I hear trains blowing for the crossing approach. Notwithstanding relatively good sighting (which could be improved by felling a fair amount of lineside jungle) and low linespeed (50mph) the crossing is listed as 'High Risk' by NR.  On one side of the railway the landowners property extends right up to a very lightly used lane which crosses the railway by an overbridge only a few hundred yards from the crossing.  On the other side of the line he could easily afford to buy the small piece of land between his property and the bridle road - so at the cost of a couple of gates, a small land purchase and some minor earthworks the crossing could be eliminated - the cost would no doubt be considerably less than the amount of money the landowner spent last year on creating his second polo pitch and it would certainly be massively less than the value of any one of the polo ponies he takes across the crossing.  Totally different circumstances of use and closure possibility than a public level crossing and more typical (apart from the polo ponies) of the situation with many occupation etc crossings.

 

When HSTs were first introduced several dozen such crossing on the GWML were closed, literally, overnight following a careful study of their use and ready alternatives and none of the closures involved a need for bridgeworks.  Separate problem from level crossings and often much simpler solutions are readily available if people care to look.

 

Trust me Mike, closure is considered for all crossings. It is simply not an option suitable for all. https://www.networkrail.co.uk/feeds/a-7-5m-upgrade-scheme-in-mid-wales-improving-rail-safety-will-begin-in-the-new-year/ .

[martin_wynne]

I don't see how that creates any extra safety

 

Well it responds to the situation in the RAIB news report, where seemingly a user opened the gates using the power button without first contacting the signaller. This would prevent a user doing that. The signaller is effectively giving a release to open the gates, without which they won't work.

 

But it does that without adding any extra connections into the signalling system between the signalbox and the crossing.

 

The extra cost is minimal. A small processor in the push button unit (a Raspberry Pi would do the job easily, cost £20) plus a weatherproof industrial number pad, cost £26:

 

 

See: http://uk.rs-online.com/web/p/access-control-keypads/8861670/

 

The extra work for the signaller consists of clicking the name of the crossing on his computer screen and reading out the 4-figure number which comes up, while talking on the phone to the crossing user.

 

regards,

 

Martin.