Southern Rail Franchise

[Mike Storey]

You're contradicting yourself in the same message. Apart from that, you're forgetting that each and every 'driver-less' system is effectively isolated from the real world. No crossings with unpredictable humans (road, pedestrian, platform) other then when the system allows for it (full height barriers at platforms with doors that open when the train doors adjacent open). It has to be this way, no system can cope with the infinite number of variables otherwise, as mentioned earlier.

 

Britain will never adopt ERTMS. First, it's "not invented 'ere", second, it is invented by that effin' EU, so no brexiteer will ever want anything to do with it. Third, the UK rail infrastructure simply isn't suited for it, there's too much Victorian technology in daily use, requiring a level of investment no UK Gov't will ever make. The Tories won't as they're anti-rail anyway and any money for what they've left of the railways needs to come from that same railways and their customers and that's simply not realistic. Labour, should they ever become dominant electable again, will not provide money that ultimately makes jobs redundant

 

As for ERMTS itself: Level 3, which is still in draft, specifies a feature known as 'gliding blocks', this is a prerequisite for efficient automation. But it's not enough. I suspect level 4 will introduce "driver managed automation" while Level 5 will upgrade that to "remote managed automation" and finally Level 6, autonomous automatic operation. Level 4 is not even being launched, let alone discussed for parametrising design conditions

 

Still, some nice tricks can be seen today, provided the right tools are in place:

 

(in German, obviously)

 

Transition PZB->LZB (German, again)

 

 

LZB test (mandatory test before driving off) ÖBB 1016:

 

 

Cab ride with transition into ETCS Level 2:

 

(starts at 6:30, in German again)

 

Sorry, but whilst often I share your sentiments and enthusiasm, this is clearly utter bo11ocks. ERTMS Level 1 has been working, not just in "test mode", on the Cambrian lines for several years now and considerable lessons have been learned. Britain was actually the first country within the UIC to go live (as opposed to test mode only) for this, although only by months.

 

HS1 has been using a variant of ERTMS since the completion of the high speed sections to St Pancras.

 

HS2 will use ERTMS, probably Level 3, but who knows in 10 years?

 

A form of ERTMS will shortly be in use through the Thameslink tunnel section, without which the 24 tph would be impossible. Crossrail will be using something similar - others will no doubt explain the technical differences. Rolling block is being advocated as long term strategy (especially by Ms Shaw in her report), as a cheaper solution for capacity resolution, even though many have debunked this as unrealistic. I don't know who is right, as my luddite tendencies have been proven wrong so many times, but the key point is that "not invented here" is a strange accusation for a network that has adopted foreign solutions as "best" over so many decades, and is unlikely to be swayed by the Brexit factor, given there are so few (none?) domestically owned providers of alternatives. Look to the railways of my Gallic hosts, or even the legally enthusiastic Deutsche Bahn AG, to see real chauvinism at work. In the Netherlands, you perhaps see Brexit as a British mindset - it is not. It is a politically populist protest by a slim majority against.....the argument about whether they chose the right target rages on, and on, and on.....

 

The UIC long precedes the existence of the EU, and the UK will, I presume, continue to be a member. Next to naff all to do with the EU really. I think you are confusing the technology with the standards imposed by TSI's, from which we may well seek to withdraw, especially given the emerging farce of non-evidence based, increased compliance clearances for 25Kv OLE, wholly unsuited to the barmy, tiny loading gauge bequeathed by our Victorian beneficiaries (demi-French GW engineer excepted, but even he built his bridges too low - didn't he have the foresight to understand that Faraday was on to something??).

[Mike Storey]

Your fares wont go down and neither would the subsidy, what would happen is that the franchisee would see a bit more profit, not as much as the supposed 'savings' from DOO because of the increased risk of an incident leading to a Court case but by the time you factor in the costs of the Court case and compensation payout for somebody losing their legs etc there is still a saving in real terms.

 

Can you show me where we have run 12 coach DOO trains with unstaffed rural stations for years please?

 

Can anyone explain to me why the McNulty report had to be deferred and rewritten when there was a change of Government?

Surely the report would have come to the same conclusion regardless of what colour Government we have!

 

For at least 10 years, Sandy, St Neots and Biggleswade for starters, any train after 19.00 weekdays, and every Saturday and Sunday after the early turn shuts up shop. In reality, more than this, as staff rarely turned up to platforms outside the peaks, from my observations. Huntingdon too, outside the peaks in reality.

 

If I had greater detailed knowledge, I suspect the same could be described for many stations on the GE south-west of Colchester and north-east of Stratford. I personally know of the inconsistent attendance of train despatch staff at Ilford, Romford and Rainham, outside the peaks in particular. Not exactly rural, but then not deemed particularly useful by drivers, it would seem. The lesson here is that there is a perceived need during peak boarding times, but otherwise not really. That suggests a compromise that all parties should be discussing?

[david.hill64]

Comments about UK not adopting ETCS signalling are wholly misguided: while we are a member of the EU we are prohibited from even developing any form of alternative, and even after we leave it is a fair bet that we would want to continue to follow the EU standards as non-EU countries already do. After all they are designed to give the rail industry a choice of suppliers for a common product with good operational and safety characteristics and that should help contain costs.

 

ETCS level 2 is being rolled out now in the UK as others have noted. As I have posted here before, when we were developing TPWS over 20 years ago this was seen as a short term (5-10 year) solution pending the roll out of ETCS level 3. This is the moving block version with no lineside signals. Railtrack bet the farm on it being available for the WCML upgrade. 23 years later there isn't a finally agreed final specification for level 3 (but it is surely imminent) and as far as I am aware the only main line railway to use something very similar is I think Kazakhstan where Bombardier has supplied a version using TETRA radio rather then the EU approved GSM-R systems. (There are trial versions in Sweden and elsewhere, but not full commercial operation).

 

Both level 2 and level 3 have the facility to become driverless and I understand that the plan is to use ATO control on Thameslink through the central section where more precise control than that possible with a driver is required.

 

The ETCS hardware also forms the basis of the latest moving block metro signalling solution used by Bombardier and in operation for example on the Bangkok Purple Line.

 

So there is no reason from a signalling control point of view why main line trains could not be driverless now. Operationally is a completely different matter though and I think we are years away from making a viable safety case for driverless operation except in self contained systems.

 

The industry is also looking at the next stage which is rolling block. In this trains would tail gate in the same way that cars do on motorways, using the braking distance of the train in front as part of the safety margin. I foresee difficulties in getting safety approvals though. With moving block giving you operational headways of 90 seconds or less, it's difficult to conceive that you might require a higher frequency.

 

Sorry a bit off topic but I wanted to counter some misconceptions.

[Coryton]

You're contradicting yourself in the same message. Apart from that, you're forgetting that each and every 'driver-less' system is effectively isolated from the real world. No crossings with unpredictable humans (road, pedestrian, platform) other then when the system allows for it (full height barriers at platforms with doors that open when the train doors adjacent open). It has to be this way, no system can cope with the infinite number of variables otherwise, as mentioned earlier.

 

Britain will never adopt ERTMS. First, it's "not invented 'ere", second, it is invented by that effin' EU, so no brexiteer will ever want anything to do with it. Third, the UK rail infrastructure simply isn't suited for it, there's too much Victorian technology in daily use, requiring a level of investment no UK Gov't will ever make. The Tories won't as they're anti-rail anyway and any money for what they've left of the railways needs to come from that same railways and their customers and that's simply not realistic. Labour, should they ever become dominant electable again, will not provide money that ultimately makes jobs redundant

 

As for ERMTS itself: Level 3, which is still in draft, specifies a feature known as 'gliding blocks', this is a prerequisite for efficient automation. But it's not enough. I suspect level 4 will introduce "driver managed automation" while Level 5 will upgrade that to "remote managed automation" and finally Level 6, autonomous automatic operation. Level 4 is not even being launched, let alone discussed for parametrising design conditions

 

I'll have one more go at this and then I should probably leave it as it's a much wider subject than the topic of this thread.

 

I'm not forgetting at all that current driverless systems are self-contained (I only mentioned LIlle to make the point that completely unmanned trains are already accepted) but it seems I'm not making my point very clearly. I will try again.

 

Let's call existing systems automatic trains. These have the following characteristics: They are designed as a complete system - the train interacts electronically with the infrastructure and has no sensors to look - for example - for an obstruction on the track. They are, so far as I know, programmed in a relatively simplistic fashion - somebody has worked out all the things the train might have to do, and written an appropriate response into the program.They operate on self-contained systems (e.g. no level crossings) and there is limited variety in the rolling stock (train lengths might vary). 

 

I fully agree that extending this approach to conventional rail would require something like much higher levels of ERTMS and probably in practise isn't going to work anyway for various reasons.

 

But now consider what I shall call autonomous trains. These would use the same technology that works now for driverless cars. As with a driverless car, they would be designed to use the existing infrastructure exactly as it is. They would have arrays of sensors so they could use existing lineside signals, brake for obstructions (e.g. a car at a level crossing), sound the horn for a passenger too close to the edge at a platform etc. Rather than being programmed to drive, they would use machine learning by "watching" what human drivers do. I suspect that if you asked the driver of a freight train how they cope with the varying characteristics of their trains they couldn't tell you in the detail you'd need to program a computer, but you don't have to. The computer itself can learn from them. I believe that the existence of self-driving cars means that autonomous trains are entirely practical.

 

Now, it is quite possible that as the market for these trains is far smaller than for cars and the benefits much less obvious it's not worth putting the effort in to developing them. It's also possible that the consequences of passing a signal at danger are so high that we'd want cab signalling (though human drivers make mistakes and you could run a system in parallel with a human driver and count how many times the train would have passed a signal at danger to prove its safety). But cab signalling is coming in whether it's ERTMS or not.

 

But so far the arguments people are coming up with here are either against automatic, not autonomous trains, or claiming that there are too many variables for it to work - and surely the existence of self-driving cars disproves that one? As does the fact that a computer has finally beaten the best Go players - the number of variables there are almost inconceivable.

 

And since writing my previous posts I've discovered that autonomous trains are being developed now - search for Rio Tinto driverless trains. It's not going smoothly whereas they already have driverless trucks, though I think the current problem is lack of mobile phone coverage and in any case the amount of development effort that's gone into this must be a fraction of that for road vehicles.

 

Anyway - the relevance of all this to Southern is that the argument is being used that DOO is an inevitable consequence of modern technology (which I don't really buy - the only technological aspect is that putting the cameras on the train makes it a bit cheaper to have the appropriate infrastructure). But looking at the longer term, I think technology is far more likely to make the driver obsolete than the person (whatever you call them) who is in the train with the passengers, not in the cab.

 

Of course looking even more widely, the existence of driverless cars and buses, and the ability to convoy traffic might threaten the existence of large chunks of the rail network.

[Derekl]

I did say 'operational safety incidents' - i.e. incidents affecting the safety of the line/train, and I know of none; operationally DOO(P) has a good safety record.

 

As far as incidents involving passenger safety are concerned the only valid comparison is between those where the decision to close the doors and depart is solely in the hands of the driver (and the on-train equipment has no design faults) and those where such decisions and actions are in the hands other members of staff be they a Guard or platform staff.  And - again - I'm not aware of any comparative statistics being published but they might lurk somewhere within an ORR Inspectorate Report because such incidents have very definitely occurred.

 

Equally - again as far as I'm aware - no summary is available of injuries or fatalities sustained by Guards when despatching their own train but I do know that such things have occurred although they seem now to be less common than they once were (usually such incidents occur when the Guard has been on the platform checking the train or watching for the tip from platform staff and has tried to join the train as/after it started to move).

 

Sorry - I don't follow this argument - it seems to me that if a passenger is injured as a result of a train departing under DOO control, albeit without damage to the line or train, there is an apparent safety issue, which needs to be addressed: the fact that the line or train is not damaged (unsurprisingly) does not seem relevant. I do agree that a comparison between passenger injuries as a result of dispatch issues under DOO or guard control would be relevant and interesting - it is possible to identify those in the RAIB reports, although it is time consuming - I noticed a few when identifying the above: there are not many over the RAIB reporting history.

 

The fact is there have been DOO dispatch injuries, some serious: those really need to be addressed before the system can be said to be "safe".

[david.hill64]

The comments about autonomous trains are interesting.

 

Work is going on to consider how to do away with manual signalling centres completely: trains will talk to each other to determine who is going to get to a conflict point first and act accordingly to set the route: ie points commanded by the train. 

 

On hearing this my immediate reaction was that it cannot happen, but given that it already is happening with driverless cars, the relatively simple case of fixed guideways ought to be quite easy. I don't want to write the safety case though!

[The Stationmaster]

Sorry - I don't follow this argument - it seems to me that if a passenger is injured as a result of a train departing under DOO control, albeit without damage to the line or train, there is an apparent safety issue, which needs to be addressed: the fact that the line or train is not damaged (unsurprisingly) does not seem relevant. I do agree that a comparison between passenger injuries as a result of dispatch issues under DOO or guard control would be relevant and interesting - it is possible to identify those in the RAIB reports, although it is time consuming - I noticed a few when identifying the above: there are not many over the RAIB reporting history.

 

The fact is there have been DOO dispatch injuries, some serious: those really need to be addressed before the system can be said to be "safe".

 

The key element with all BR versions of DOO right from the start was operational safety - in other words would operation of a train by a  single person (the Driver) make the railway any less safe than having two people on the train.  And the answer is that it hasn't made it any less safe.

 

As far as entraining/alighting passengers are concerned DOO(P) has to conform with various standards which make that process as safe as possible and they are pre-reqisuite for rolling stock and stations.  Thus for example it is quite permissible under the original DOO(P) conditions to operate a train with slam door coaching stock and no central door locking - and it was done (Slough to Windsor branch) but it required staffed station platforms and additional signals operated by platform staff to advise the Driver to stop the train if there was an open door etc.  It worked without any problems, and no incidents of any kind,for the whole remaining life (several years) of slam door trains.

 

Equally at some stations on the GWML in the London suburban area, and beyond, passenger trains are worked DOO at stations which have neither mirrors not monitoring cameras - the only incident I'm aware of is that at Hayes (which revealed a potential problem in the door closing sensor system).  Passenger injury incidents were changed in the overall reporting mechanism (about 20+ years ago) so there should already be more than sufficient data to indicate whether or not that part if the system is 'safe' (depending of course on what is regarded as 'safe') compared with other methods of train despatch - but I've still not seen a summary.

[phil-b259]

Sorry - I don't follow this argument - it seems to me that if a passenger is injured as a result of a train departing under DOO control, albeit without damage to the line or train, there is an apparent safety issue, which needs to be addressed: the fact that the line or train is not damaged (unsurprisingly) does not seem relevant. I do agree that a comparison between passenger injuries as a result of dispatch issues under DOO or guard control would be relevant and interesting - it is possible to identify those in the RAIB reports, although it is time consuming - I noticed a few when identifying the above: there are not many over the RAIB reporting history.

 

The fact is there have been DOO dispatch injuries, some serious: those really need to be addressed before the system can be said to be "safe".

 

Risk is all around us and as with the level crossing thread I remind people we have a legal principle in this country - fully tested by the courts - that when it comes to assessing risks of existing or new practices the "As low As Is Reasonably Practicable" applies.

 

Thus just because 'some' DOO incidents have occurred in no way makes the practice 'unsafe' as a principle. Yes there may be individual elements that could be improved - however hindsight is a wonderful thing. As the RAIB state they key is that when incidents occur we learn from them. The Hayes incident showed an issue with door seals / interlock / obstruction detection, and the industry has respond to that risk once it was bought to light.

 

The Purley smash in 1989 showed up a risk that the industry / Government didn't learn from - hence we got Labroke Grove many years later before TPWS was introduced

 

The Belgrove smash in 1989 showed up a risk having guards give the RA to drivers that took some time to be addressed (via the DRA button in the cab)

 

The death of a passenger in Liverpool last year (which I believe is the subject of a criminal prosecution) shows that having a Guard is no guarantee of safety during dispatch.

[Oldddudders]

The Inner Suburban part of the Southern Franchise has been DOO for about 25 years. I wrote the Investment Submission for the works in the late '80s. To the best of my knowledge on-train safety issues affecting passenger personal security have not been significant - and some of the areas through which it runs are potentially rather less savoury than the nicer towns of East and West Sussex, with a higher risk of obnoxious characters.

 

That leaves the issue of train despatch. As Stationmaster has pointed out, there are clear requirements for the safe operation of trains, and if these have been followed I see no reason why 12-car DOO isn't just as safe as any shorter formation. The suburban area has curved platforms aplenty, so actually the length of the train is not significant. As long as the driver has adequate visual coverage of the entire train, safety is assured.

 

Where DOO is not safe is where the driver cannot start the train immediately on satisfying himself all is well. The suburban services from Euston to Northampton were converted to DOO in early 1990, but the scheme was stopped within days. The brand new 321 units had had the driver's seat moved away from the nearside droplight towards the middle of the cab, a modification intended to improve his view ahead, but thus requiring him to close the droplight and move to his seat before starting away. The shore equipment was never used again. 

 

And safe despatch is not guaranteed by having on-train and/or on-platform staff. Even in the "good old days" drivers were known to leave without a proper signal from the guard - or even a proceed aspect! - and shortly before I became a station manager in 1979 there had been a distressing accident at my station involving a passenger falling beneath a moving slam-door train. The platform staff member who might have prevented this was so traumatised by his potential part in the incident that he never worked again. 

[Glorious NSE]

Anyway - the relevance of all this to Southern is that the argument is being used that DOO is an inevitable consequence of modern technology (which I don't really buy - the only technological aspect is that putting the cameras on the train makes it a bit cheaper to have the appropriate infrastructure). But looking at the longer term, I think technology is far more likely to make the driver obsolete than the person (whatever you call them) who is in the train with the passengers, not in the cab.

 

From this passenger's point of view, i'd be fine with driverless autonomous trains with onboard staff (customer host/safety backup) coming in at some currently undetermined point in the future, but I am strongly against a 1980s style cost-cutting approach with no customer facing staff on stations or trains though (and the resulting increase in crime and antisocial behaviour...) - been there, done that.

 

Overground has shown what a transformation you can make if you staff a system with the intent on making a difference to the travelling public, that approach does come at a cost.  

[phil-b259]

Overground has shown what a transformation you can make if you staff a system with the intent on making a difference to the travelling public, that approach does come at a cost.  

 

Oh indeed - but the Overground has chosen to staff stations - NOT continue with Guards.

 

Based on some peoples views this is apparently unacceptable - despite it actually being a very good way of addressing 99% of the issues that DOO brings with it on a daily basis.

 

Of course part of the reason Overground went for this mode of perpetration is that station staff (even ones qualified to dispatch) are on lower wages than Guards - so it actually works out as pretty revenue neutral overall even if the numbers of staff employed go up to facilitate it.

 

Its also more easy to manage staff shortages as staff can be more easily moved around to cover gaps than Guards with their route knowledge requirement.

 

Up thread there was some talk of GTR being removed from the franchise - this BBC article may be of interest, and to the organisations credit they do make pains to pint out that simply replacing the top management will do nothing to solve most of Southern's woes - particularly if said new management are duty bound to do what the DfT tell them.

 

http://www.bbc.co.uk/news/business-38771216

Edited January 29, 2017 by phil-b259

[jjb1970]

There is a lot of truth in the old adage of nothing looking as old fashioned as yesterday's vision of today.

 

Autonomy is a very happening area of technology, military agencies have been pumping money into developing autonomous systems for years, and I mean autonomous and not remotely piloted vehicles. Driverless cars will be autonmous and are not far from making the leap from limited test programs to going on sale (or availability to lease/hire). I did some work involving autonomous underwater vehicles for military use, not in technical terms but advising on regulatory requirements and gaps.

 

I think introducing autonomy to a legacy system would be very challenging. In some ways an autonomous train would pose fewer challenges than an aerial vehicle or car in terms of the vehicle itself but the wider system would be problematic. An autonomous rail system would be just that, a system, not an autonomous train. Short of taking the system out of service to re-boot and then go live as a fully autonomous system I could see all sorts of pitfalls converting a legacy railway to autonomous operation.

 

For all that, I do think we will see a steady evolutionary move towards driverless trains in some form.

[Coryton]

Risk is all around us and as with the level crossing thread I remind people we have a legal principle in this country - fully tested by the courts - that when it comes to assessing risks of existing or new practices the "As low As Is Reasonably Practicable" applies.

 

And in some cases that risk can be rather high indeed, an example being bus stops on 70 mph dual carriageway roads where passengers have to cross the road on foot.

 

Here "reasonably practical" means bunging up some signs for drivers warning that pedestrians might be in the road.

[Mike Storey]

The comments about autonomous trains are interesting.

 

Work is going on to consider how to do away with manual signalling centres completely: trains will talk to each other to determine who is going to get to a conflict point first and act accordingly to set the route: ie points commanded by the train. 

 

On hearing this my immediate reaction was that it cannot happen, but given that it already is happening with driverless cars, the relatively simple case of fixed guideways ought to be quite easy. I don't want to write the safety case though!

 

Not just "going on" but active with the overlays now being trialled for more advanced versions of ARS at certain ROC's, by Hitachi I think.  Not quite "trains talking to each other", but the ROC computer logic using more than just the planned TT to suggest changes to the plan to the signaller. The next stage has the system making those decisions and implementing them, with the signaller able to intervene if he/she sees a problem with that decision. I gather the Thameslink core tunnel section will work this way from 2018, basing its decisions on realtime performance of services approaching it from each end. I take my info from reading, then re-reading the Roger Ford descriptions and the IRS summaries, so stand to be corrected on the detail. Could be the ultimate (on current technology) test of the business case for such developments. CrossRail will be watching with interest, albeit its planned system is different.

 

On driver-less trains on the mixed traffic, main line, I think we are a generation or two away from that. Mainly due to the high costs of implementing the technology, not the technology itself. Just look at G3 coverage to understand that, apart from the fact that it must be safety critical. Line-side beacons had to be the solution for HS1 (and almost all of high speed lines in France, Spain, Germany and Italy, and I would guess elsewhere). If Balises did not work for the WCML, where on earth will they work? The new EuroSat GPS, which is more accurate than the US military system it aspires to replace, may take it forward more quickly, but it still has huge holes. Can we access that now?......

[Glorious NSE]

Based on some peoples views this is apparently unacceptable - despite it actually being a very good way of addressing 99% of the issues that DOO brings with it on a daily basis.

 

Certainly for urban networks with fairly high frequencies which are relatively busy and have closely spaced stations it's a great model.

 

I'm not convinced a 'one size fits all' approach is a good idea for instance - having staff on hand at stations is fine when distances are very short and frequency is high, but you wouldn't want to apply the same principle to (for instance) an Intercity service where the next station call may be an hour away, or to a rural route with very infrequent services and low customer numbers - on both of those having a member of staff on the train may well be the best option.

 

Which is why I keep wishing (as I do with all other 'customer quality' issues on todays railway) that the DfT was taking a lead on specifying a minimum spec for ALL operators to live up to, rather than trying to drive an 80s style budget lowest price at all costs railway. 

[Derekl]

"The Stationmaster: The key element with all BR versions of DOO right from the start was operational safety - in other words would operation of a train by a single person (the Driver) make the railway any less safe than having two people on the train.  And the answer is that it hasn't made it any less safe."

 

Err, sorry, evidence for that assertion. I have just provided 4 examples of injuries of varying significance quite clearly under DOO operation -I could find few involving injuries on guard's dispatch.If you want to assert that DOO is "any less safe" you need to provide statistical evidence, and clearly from your own assertions it ought to be available: 

 

"Passenger injury incidents were changed in the overall reporting mechanism (about 20+ years ago) so there should already be more than sufficient data to indicate whether or not that part if the system is 'safe' (depending of course on what is regarded as 'safe') compared with other methods of train despatch - but I've still not seen a summary."

 

I am not a great enthusiast for conspiracy theories, but I do wonder why not, possibly because it does not look too good?

 

"Phil-b259: Risk is all around us and as with the level crossing thread I remind people we have a legal principle in this country - fully tested by the courts - that when it comes to assessing risks of existing or new practices the "As low As Is Reasonably Practicable" applies."

 

I assume you are reminding me of "legal principle" of "as far as reasonably practicable" - in fact it is not an issue for testing by the courts or a legal principle, but a frequent inclusion in regulation and statutory instrument as the standard to be complied with. The balance is the risk of injury and the potential seriousness of the injury. The problem with this issue is that the potential for serious injury is high.

 

And: "Thus just because 'some' DOO incidents have occurred in no way makes the practice 'unsafe' as a principle."

 

Err, sorry again, but why not? If at least 4 people have suffered serious injuries on train dispatch under DOO: should one not be asking a few questions, rather than just saying this was signed off as safe?

 

And: "Thus just because 'some' DOO incidents have occurred in no way makes the practice 'unsafe' as a principle. Yes there may be individual elements that could be improved - however hindsight is a wonderful thing"

 

Yes, well hindsight is wonderful, but that is not where we are - there have  been DOO incidents that have occurred, so we are not talking about hindsight, or what might have happened, these did. The RAIB does not suggest or address in any of these incidents whether DOO of itself was a risk element, a serious failure in the reporting and investigation.

 

And: "The Hayes incident showed an issue with door seals / interlock / obstruction detection, and the industry has respond to that risk once it was bought to light."

 

Has it: have door seals been modified over a number of fleets?

 

"Olddudders: “Even in the "good old days" drivers were known to leave without a proper signal from the guard - or even a proceed aspect! - and shortly before I became a station manager in 1979 there had been a distressing accident at my station involving a passenger falling beneath a moving slam-door train."

 

Sure, but drivers leaving without signal is not really the issue - was the 1979 incident the one at Burgess Hill, in Sussex?

[great central]

Another thing to bear in mind is that we know there have been incidents under both guard and DOO despatch.

What we don't know, and unfortunately I doubt ever will, is how many 'near misses' there have been under the relative systems. I can think of a good many instances personally over a reasonably short career where, without another human present either on train or platform to intervene, there would have been the potential for death or serious injury. I think all my colleagues would have a similar catalogue of people having to be saved from their own foolishness. 

[phil-b259]

"The Stationmaster: The key element with all BR versions of DOO right from the start was operational safety - in other words would operation of a train by a single person (the Driver) make the railway any less safe than having two people on the train.  And the answer is that it hasn't made it any less safe."

 

Err, sorry, evidence for that assertion. I have just provided 4 examples of injuries of varying significance quite clearly under DOO operation -I could find few involving injuries on guard's dispatch.If you want to assert that DOO is "any less safe" you need to provide statistical evidence, and clearly from your own assertions it ought to be available: 

 

"Passenger injury incidents were changed in the overall reporting mechanism (about 20+ years ago) so there should already be more than sufficient data to indicate whether or not that part if the system is 'safe' (depending of course on what is regarded as 'safe') compared with other methods of train despatch - but I've still not seen a summary."

 

I am not a great enthusiast for conspiracy theories, but I do wonder why not, possibly because it does not look too good?

 

"Phil-b259: Risk is all around us and as with the level crossing thread I remind people we have a legal principle in this country - fully tested by the courts - that when it comes to assessing risks of existing or new practices the "As low As Is Reasonably Practicable" applies."

 

I assume you are reminding me of "legal principle" of "as far as reasonably practicable" - in fact it is not an issue for testing by the courts or a legal principle, but a frequent inclusion in regulation and statutory instrument as the standard to be complied with. The balance is the risk of injury and the potential seriousness of the injury. The problem with this issue is that the potential for serious injury is high.

 

And: "Thus just because 'some' DOO incidents have occurred in no way makes the practice 'unsafe' as a principle."

 

Err, sorry again, but why not? If at least 4 people have suffered serious injuries on train dispatch under DOO: should one not be asking a few questions, rather than just saying this was signed off as safe?

 

And: "Thus just because 'some' DOO incidents have occurred in no way makes the practice 'unsafe' as a principle. Yes there may be individual elements that could be improved - however hindsight is a wonderful thing"

 

Yes, well hindsight is wonderful, but that is not where we are - there have  been DOO incidents that have occurred, so we are not talking about hindsight, or what might have happened, these did. The RAIB does not suggest or address in any of these incidents whether DOO of itself was a risk element, a serious failure in the reporting and investigation.

 

And: "The Hayes incident showed an issue with door seals / interlock / obstruction detection, and the industry has respond to that risk once it was bought to light."

 

Has it: have door seals been modified over a number of fleets?

 

"Olddudders: “Even in the "good old days" drivers were known to leave without a proper signal from the guard - or even a proceed aspect! - and shortly before I became a station manager in 1979 there had been a distressing accident at my station involving a passenger falling beneath a moving slam-door train."

 

Sure, but drivers leaving without signal is not really the issue - was the 1979 incident the one at Burgess Hill, in Sussex?

 

Let me remind you that the courts have found that in lots of cases it is not necessary for an organisation to prevent deaths at any price. You might not like it - but the simple fact is that deaths are actually considered acceptable in this country.

 

All that the Courts and the HSE, etc demand is that (i) The risks have been correctly assessed and (ii) The mitigation strikes the right balance (as expressed in a mathematical ratio) between cost of the measures and the number of lives that may be saved - too few lives saved or too grater cost are valid reasons for not doing anything under the ALARP methodology we employ in this country from building sites - to railways - to the NHS.

 

If we went down you route of absolute safety we would fit all signals with TPWS - because the risk of death or serious injury if one train rams into the back of another is just as bad as something crashing into a vehicle on a level crossing. Fortunately more enlightened councils recognise that doing so is a waste of money that could be better spent on other things and the risk / lived saved ratios doesn't come close to the accepted thresholds for such an action to be justified so auto signal on plain line generally lack TPWS protection.

 

Similarly with DOO you are living in cloud cuckoo land if you think that the amount of 'serious incidents' shows any significant statistical variation with non DOO services. Again the safety regulators and courts take a more sensible view that provided the correct technical mitigation is in place then the risk to passenger is below that magic threshold at which we say things are dangerous. We have 30 years of experience of DOO now and the statistics don't lie - unless you chose to be deliberately selective about them. Furthermore you also seem blind to a number of serious incidents (like the Belgrove smash in 1989 where the Guard was directly responsible for starting the chain of events by giving the driver the RA while the signal was at red) where the presence of a Guard has either made absolutely no difference to the outcome or made things worse by not intervening when they could / should.  That is why the RAIB don't question DOO as a principle - because unlike you they actually look at the evidence of the past 30 years and can find no fault with any of the principles behind it - even if occasional upgrades (such as better door edge sensors) should be fitted post the Hayes incident.

Edited January 30, 2017 by phil-b259

[Oldddudders]

"

 

"Olddudders: “Even in the "good old days" drivers were known to leave without a proper signal from the guard - or even a proceed aspect! - and shortly before I became a station manager in 1979 there had been a distressing accident at my station involving a passenger falling beneath a moving slam-door train."

 

Sure, but drivers leaving without signal is not really the issue - was the 1979 incident the one at Burgess Hill, in Sussex?

No, it was not Burgess Hill. And I cannot see any difference, from the passenger's point of view, between a train departing because the DOO driver has satisfied himself it is safe to do so when it is not, and a non-DOO driver proceeding without a right-away from guard or platform staff. The train moves when it shouldn't, and we all agree that is to be avoided. I think.

[roythebus]

Good points raised in your last paragraph Roy, and as you're no doubt aware a 1,200 ton freight train is by no means the heaviest on today's network (ours at FLHH vary anywhere between 300 and 3,300). As for braking characteristics, you can have the same loco and same set of wagons on the same route five days in a row and the braking 'feel' will be different each time.

 

Yes, I was aware that there has been a remarkable increase in freight train weight since my days driving on BR. the heaviest we worked at Rugby was about 1200 tons of Freightliners from Willesden to Crewe. That took some shifting, especially with an 87 on a damp rail. No amount of automation could deal with that in my view. 

[roythebus]

Having been following this thread at a distance and made a couple of contributions. it seems to have gone a long way from the original point! There's been a number of valid points made by both professional (and former) railway staff, some of whom have had personal experience of the actual situation and others who have a good idea of what needs to happen.

 

Most of the railway staff have contacts with fellow railwaymen in other countries we can compare notes with, and us former staff still stay in touch with events.

 

The late Alec Swain (former high-up on the BR hierarchy and chairman of the Model Railway Club) had various chats with me when he was on the BR payroll, when the Bed-Pan project was being undertaken in the early 1980s. He gave me a cartoon which showed the "moving block" system proposed for the St.Pancras system which, when i find it, will post on here.

 

I'd suggest that the "moving block" was probably accidentally invented with the introduction of the 313/508/455 stock when the braking systems left a lot to be desired. As Rugd says, every train handled differently, making any form of automation difficult. However, the dispute isn't over automation. It is about passenger safety. While the GM may have 12 cars with no platform dispatch staff OUTSIDE the peaks, that implies it has platform staff during the peaks. I don't suppose there would be the same high level of risk on the Southern outside the peaks. What has to be borne in mind is that every "passenger incident" causes delay not only to the train concerned, but on a network like the Southern with trains at 2 minute headways, causes considerable disruption to a large number of people in a short period of time.

 

There have always been boarding and alighting incidents whether or not there are guards on the trains, they will always happens because, as I've aid before, you cant legislate for stupid. This applies to the idiots at the DaFT and their political master who come up with the crazy ideas that disrupt so many people's lives. What Dutch Master says about the NS system on train staffing seems a sensible idea to adopt for better passenger safety.

 

As for the cost of a life, does anyone remember the work done on a cost/benefit analysis following the Ufton Nervett (I think) crash where an HST hit a car? The cost of fitting seat belts to all trains of all things was found to be horrendous for the amount of lives the proposal would save. Luckily since then there have been hardly any deaths caused by train crashes (maybe none, I don't have the stats to hand) . No matter what happens, there will ALWAYS be boarding incidents  with the current crap equipment available to drivers, (and there's been some excellent pictures posted on a couple of railway staff FB groups of the platform monitors in rain and in-cab monitors showing that the system is not fit for purpose).

[The Stationmaster]

"The Stationmaster: The key element with all BR versions of DOO right from the start was operational safety - in other words would operation of a train by a single person (the Driver) make the railway any less safe than having two people on the train.  And the answer is that it hasn't made it any less safe."

 

Err, sorry, evidence for that assertion. I have just provided 4 examples of injuries of varying significance quite clearly under DOO operation -I could find few involving injuries on guard's dispatch.If you want to assert that DOO is "any less safe" you need to provide statistical evidence, and clearly from your own assertions it ought to be available: 

 

The evidence is that there has not been a single case of any sort of operational safety incident where DOO (of any sort) was reported as an additional contributory factor to that incident - simples.  If you don't believe me you can plough back through 30 years of Reports and annual summaries to find the ones I and the railway industry might have missed.   Although if you believe in conspiracy theories you could perhaps say that as the conditions for the introduction of DOO were jointly produced between BR and the railway Inspectorate they'd hardly be likely to show themselves up in a bad light.

 

Simple fact DOO is operationally as safe as operating a train with a Guard or other second crew member.

 

You provided evidence - sorry a list - of incidents relating to passengers joining or attempting to join trains.  That list was not balanced by one listing similar incidents on trains which had Guards (at least one of which involved a passenger fatality - from which bald statistic you could as readily conclude the equally baseless assertion that having a Guard in charge of closing train doors is more likely to kill people than not having a Guard because that is what the statistics 'prove')

[roythebus]

I have to agree with you there Mike, I was going to raise the matter of the comparison between a train with a guard and the train without!

[royaloak]

The death of a passenger in Liverpool last year (which I believe is the subject of a criminal prosecution) shows that having a Guard is no guarantee of safety during dispatch.

Is that the incident where the elderly lady tried to get on when the doors were closed and fell between the gap?

 

If so the train never moved and the guard was cleared of any wrongdoing by MerseyRail and the RAIB (he was actually praised for the speed and professional way he dealt with it) but the Criminal Protection Service Crown Prosecution Service decided there was a case to answer under some Health and Safety legislation and tasked the BTP with gathering the evidence. 

 

That case has been adjourned 3(I think) times now and doesnt look to be heard any time soon.

 

Or is it a different example?

[phil-b259]

No matter what happens, there will ALWAYS be boarding incidents  with the current crap equipment available to drivers, (and there's been some excellent pictures posted on a couple of railway staff FB groups of the platform monitors in rain and in-cab monitors showing that the system is not fit for purpose).

 

All that shows is that with the monitors and camera combination fitted to some trains, the images are sometimes not as clear as is expected. That can be due to equipment defects or certain weather conditions - neither of which are things that cannot be fixed or that will pass in a few hours. If the images are such that the driver does not feel they are suitable then the mitigation is we get the driver to make a manual check - if necessary walking the entire length of the platform and make arrangements to have the train taken out of service - or do what already happens at busy stations - employ platform staff who can assist with dispatch if necessary.

 

The upshot being it does not prove the principle of DOO is unsafe in the slightest - merely that the technical equipment provided to mitigate the extra risks is good, reliable and suitable plans are drawn up to address when it fails.

Edited January 31, 2017 by phil-b259