Jump to content
 

Class 800 - Updates


Recommended Posts

  • RMweb Gold

Great to think a fellow RMwebber was on the front, thankyou!

 

I was indeed baffled by the time 1K08 took to actually arrive, and ditto my train. There is a nifty mimic diagram by the passenger lounge and loos on Platform 7, and it seemed to take three fortnights for trains to enter the station. In a previous life, I was aware of something similar at Paddock Wood, where a 20mph entry into the up platform loop necessitated severe approach control on the signal in a 90mph line. Every stopping train came to a stand for no reason. A change of Divisional Manager got the necessary works sorted, and everyone was happier. If GWR get together with NR, I assume the same could be achieved at Reading - and if schedules were unchanged, voila, extra recovery time!

Link to post
Share on other sites

But you can still network a W95 computer with a W10 one, or with a new or old Mac or Linux PC, or a phone, or Rasberry Pi, etc. There are standards which still apply, they have been added to but the basics remain the same. The devices may do the function differently internally, but the their external communications are the same.

The older train should be able to control all the functions on the new train that the older train has. As new functions are introduced (eg door interlock, bodyside CCTV, reservation systems, etc) these should have new standards created, overlaid on the base standards.

There is no reason why a new train shouldn’t be able to couple to a 20 year old train and be able to control the basics needed to get it in the move (at full line speed, not just talking recovery) except sheer ineptitude in planning and specification.

 

Yes, the old OS will talk to the new OS, but lots of luck getting the software running to behave in a similar manner.  Word 95 is unlikely to be able to deal with any files created by Office 365 is a reasonable manner, and there are lots of other areas where the applications can cause lots of trouble.

 

Now add in that we aren't talking standard software.  We are talking about software which is life critical - it controls the brakes! - which means a lot of stuff is done differently than your standard web browser for example.  The specification is a lot more strict, the testing is more strict, and if you don't physically test it to work with that 20 year old train the the fail safe is to refuse to work with it at all because you can't make assumptions that the 20 year old software doesn't have quirks, bugs,  or other unique things that could cause loss of life.

Link to post
Share on other sites

Yes, the old OS will talk to the new OS, but lots of luck getting the software running to behave in a similar manner.  Word 95 is unlikely to be able to deal with any files created by Office 365 is a reasonable manner, and there are lots of other areas where the applications can cause lots of trouble.

 

Now add in that we aren't talking standard software.  We are talking about software which is life critical - it controls the brakes! - which means a lot of stuff is done differently than your standard web browser for example.  The specification is a lot more strict, the testing is more strict, and if you don't physically test it to work with that 20 year old train the the fail safe is to refuse to work with it at all because you can't make assumptions that the 20 year old software doesn't have quirks, bugs,  or other unique things that could cause loss of life.

Except that the processor operating systems do not normally talk to each other directly - they do so via interfaces and databuses.

 

I'd add to that that unless things have changed very recently, it was normal to leave the ultimate control of the brakes to hard-wired circuits, just in case. The service brake (which is really just a call for negative tractive effort) may be done electronically, but the emergency brake was/is done by straight analogue circuits whose failure modes are predictable.

 

Jim

  • Like 1
Link to post
Share on other sites

  • RMweb Gold

A consequence, possibly, of track engineers, signal engineers and operators not talking to each other in a joined up manner during the design stages, with the designated project engineer being almost certainly a specialist in one discipline, not necessarily even one of the three mentioned.

 

Jim

 

I get the strong impression that the current Reading layout was designed by PerWay engineers simply in order to get everything into a set space and in some cases without any consideration at all of braking and acceleration curves for even diesel trains let alone electrics.  The Relief Lines side of the station is probably the worst example of this but the approach to what is now Platform 7 off the Down Main is another as is some of the geometry of the pointwork at both ends of the Westbury diveunder lines.  I have a nasty feeling that the consequences of the shortcomings in the design are going to come home to roost in ongoing maintenance bills and possibly premature replacement (by normal lifetime standards) of certain pointwork, especially the ludicrously slow scissors crossover at the London end of Platform 13 & 14 which is a country mile away from a Class 387 acceleration curve.

Link to post
Share on other sites

Yes, the old OS will talk to the new OS, but lots of luck getting the software running to behave in a similar manner. Word 95 is unlikely to be able to deal with any files created by Office 365 is a reasonable manner, and there are lots of other areas where the applications can cause lots of trouble.

 

Now add in that we aren't talking standard software. We are talking about software which is life critical - it controls the brakes! - which means a lot of stuff is done differently than your standard web browser for example. The specification is a lot more strict, the testing is more strict, and if you don't physically test it to work with that 20 year old train the the fail safe is to refuse to work with it at all because you can't make assumptions that the 20 year old software doesn't have quirks, bugs, or other unique things that could cause loss of life.

To continue the analogy, your new train doesn’t save its files meant to be read by the old train as .docx , it saves them as .txt* Docx might be able to contain the formatting frills like automated anouncements and passenger loading info but the standards for critical things like braking were set in a much older format that both trains can understand.

*ignoring for a moment that there are various text file formats that differ slightly!

Link to post
Share on other sites

As technology moves on, so too do the interface protocols. It might be possible to keep such things static, but it may also be that the 387 requires information to be shared within a train formation that a 375 cannot provide, safely interpret, or even simply pass along (in the case of a 387-375-387 formation).

 

Usually engineers would strive to achieve interoperability, particularly where they have the details of what they're interfacing with, but it's not always possible without severely compromising the operation of the new device.

Link to post
Share on other sites

  • RMweb Premium

Interface management and agreeing communication protocols is a mature concept and many industries and applications manage communication between complex systems (many safety critical) despite the hardware often being sourced from different manufacturers and/or owned and operated by different entities.

Link to post
Share on other sites

  • RMweb Premium

It is all very well saying systems should be able to 'talk' to each other. A former employer of mine did an IT audit and found they had 75 different systems with many duplicated databases - none of them spoke to each other, all the data transfer between systems was done manually, in one case I was having to screen scrape data for a time critical valuation process where the deadline was just 4 hours into the first working day of each month.

Link to post
Share on other sites

Is it possible there are two separate conversations going on on this thread?

Quite possibly - one by those who have a reasonable, but not complete, working knowledge of how train control systems, the other by those with much less knowledge.

 

Few, if any amongst us have detailed knowledge of the various train's systems (and if they did, they may not be able to speak about them), and the concept that different trains should be able to be compatible with each other, at least for rescue purposes, seems to be one that has gone the way of most joined up thinking in the industry. The days of the railway system operator taking the lead as a guiding mind have been subsumed by the salesmen from the car builders and the relative inexperience of the train operating company managements.

 

Jim

 

Jim

  • Like 2
Link to post
Share on other sites

  • RMweb Gold

Quite possibly - one by those who have a reasonable, but not complete, working knowledge of how train control systems, the other by those with much less knowledge.

 

Few, if any amongst us have detailed knowledge of the various train's systems (and if they did, they may not be able to speak about them), and the concept that different trains should be able to be compatible with each other, at least for rescue purposes, seems to be one that has gone the way of most joined up thinking in the industry. The days of the railway system operator taking the lead as a guiding mind have been subsumed by the salesmen from the car builders and the relative inexperience of the train operating company managements.

 

Jim

 

Jim

 

To be fair Jim it isn't just a symptom of the privatised railway but we were already on that road in BR days.  Assisting a failed HST was hardly a simple task, especially in sheer physical terms, while having once tried it in ideal conditions coupling an HST to a failed (for exercise purposes) 15X dmu was a near impossibility and in the end we gave up.   Now the latter was in a far from everyday situation with not only broad daylight on level straight track with minimal ballast shoulder but a couple of Traction Inspectors plus a Traffic Inspector and the two traincrews on hand and somebody else on hand to read the instructions on how to do it once problems were encountered.

Link to post
Share on other sites

  • RMweb Premium

I remember when 156's were relatively new and they took over a lot of the non-WCML services that operated through Carlisle. They were regularly joined and split at Carlisle and for a few months it was a lottery as to whether or not they'd couple normally, it was quite funny seeing some drivers get progressively more violent in their push up against the other unit to try and get the things to couple (though I can't imagine it did much good for the units concerned).

Link to post
Share on other sites

  • RMweb Premium

Thanks for the info about Thunderbird possibility. I reckon there is an opportunity for one or two owners of loco's to get some couplings sorted and be available to rush about dragging the things away as quickly as possible. The ECML will need several; they use DB 67s at the moment I think but I have no idea what couplings they have?

It will be interesting to read info from any Crew that are on RMW & that drive 800s if they are permitted to share their experiences of the behaviour of the Units in different circumstances.

Phil

Link to post
Share on other sites

I get the strong impression that the current Reading layout was designed by PerWay engineers simply in order to get everything into a set space and in some cases without any consideration at all of braking and acceleration curves for even diesel trains let alone electrics.  The Relief Lines side of the station is probably the worst example of this but the approach to what is now Platform 7 off the Down Main is another as is some of the geometry of the pointwork at both ends of the Westbury diveunder lines.  I have a nasty feeling that the consequences of the shortcomings in the design are going to come home to roost in ongoing maintenance bills and possibly premature replacement (by normal lifetime standards) of certain pointwork, especially the ludicrously slow scissors crossover at the London end of Platform 13 & 14 which is a country mile away from a Class 387 acceleration curve.

 

Most likely the data on braking and acceleration given to p/way design was a work of fiction. It always has been for new stock given the time lead to design s&c. I redid most of the curves on the Ely to Norwich section based on the projected performance of the Class 158s only to find that I need not have bothered on most of it given that they did not accerate anything like as fast as expected (assuming they did not take the coping stione of the platforms!).  We came to conlusion that they forgot to take account of the power for a/c and other aux.

Link to post
Share on other sites

  • RMweb Premium

A consequence, possibly, of track engineers, signal engineers and operators not talking to each other in a joined up manner during the design stages, with the designated project engineer being almost certainly a specialist in one discipline, not necessarily even one of the three mentioned.

 

Jim

 

Or the project engineer simply being nervous of being scapegoated if a driver makes a mistake and ends up SPADing the signal. Granted it might take years to happen but possibility is still a possibility regardless of how likely it is that it will actually happen.

 

While it would be nice to think that railway staff would always look to produce the best design engineering wise, the way everything is outsourced these days means its very easy to become the fall guy for wider procedural failings and as such caution is usually the watchword.

 

The RAIB report into the Waterloo incident is worth a read in this respect as it shows just how complicated things have become in the field of resignalling or layout alterations with a bewildering array of partnerships, consultants, contractors, testers, verifiers, etc

Link to post
Share on other sites

The RAIB report into the Waterloo incident is worth a read in this respect as it shows just how complicated things have become in the field of resignalling or layout alterations with a bewildering array of partnerships, consultants, contractors, testers, verifiers, etc

 

While I am not an advocate of wholesale rail renationalisation, I do believe that a strong case could be made to take a lot of these outsourced activities - particularly safety critical ones - back under NR direct management. I am not proposing that NR gets into signalling design, but installation and the various verification and validation works (and I say that working for a company that would lose valuable contracts if this happened).

  • Like 3
Link to post
Share on other sites

  • RMweb Gold

It's all very well saying it should be possible for different software versions to talk to each other (and they should), and the MS Word analogy is good, but the difference between that & software controlling a train or signalling system is that if MS Word crashes or fails to read a file properly, it doesn't generally have the deaths of dozens if not hundreds of people, potentially within seconds, as a consequence.

 

It's not easy to test software under all possible sets of circumstances, and where a safety critical function is involved, you have to be bang on, or else you kill people.

 

Regarding multiple interfaces & software versions, I work in the mobile radio industry. My current employer is looking at replacing an asset management system that is over 20 years old, and creaking under the strain. It's a good system, but simply not being used in the way it was intended. It has to interface with equipment vendors' software, workflow management systems, performance databases etc. Most of the data exchange between these systems is by file exchange, rather than by direct communication. They exchange files which are compiled in an agreed format. It works, maybe not as well as it could, and it does take a bit of managing when new vendors or software gets integrated, but it does work.

 

Since we don't really have a single integrated system, most rollout is done using MS Excel spreadsheets, which means multiple versions float around, all slightly different. If Mr Gates has a magic line of code in MS Excel which turned it off, a fair amount of the world would grind to a halt!

Link to post
Share on other sites

While I am not an advocate of wholesale rail renationalisation, I do believe that a strong case could be made to take a lot of these outsourced activities - particularly safety critical ones - back under NR direct management. I am not proposing that NR gets into signalling design, but installation and the various verification and validation works (and I say that working for a company that would lose valuable contracts if this happened).

I would agree, although one thing that it needs would be for the right people in NR to be able to take the necessary engineering decisions. Ever since the days of Railtrack, everything has been reduced to does it comply/not comply with standards - engineering by numbers, as it were. Taking that sort of approach, whilst at the same time trying to shift technical risk onto everyone else, ie the design consultants (who aren't in business to take risks either), results in non-joined up design.

 

Jim

  • Like 1
Link to post
Share on other sites

  • RMweb Gold

I would agree, although one thing that it needs would be for the right people in NR to be able to take the necessary engineering decisions. Ever since the days of Railtrack, everything has been reduced to does it comply/not comply with standards - engineering by numbers, as it were. Taking that sort of approach, whilst at the same time trying to shift technical risk onto everyone else, ie the design consultants (who aren't in business to take risks either), results in non-joined up design.

Jim

My observation is that in the evil Railtrack era, when doing nothing while watching the money roll in was the game, the sort of senior ex-BR engineers who could see what was wanted were expensive and regarded as unnecessary. Lots of bright, hard-working younger people took their place, but lacked their experience and foresight. NR is no doubt still struggling to recover the required balance.
Link to post
Share on other sites

While I am not an advocate of wholesale rail renationalisation, I do believe that a strong case could be made to take a lot of these outsourced activities - particularly safety critical ones - back under NR direct management. I am not proposing that NR gets into signalling design, but installation and the various verification and validation works (and I say that working for a company that would lose valuable contracts if this happened).

 

 

My observation is that in the evil Railtrack era, when doing nothing while watching the money roll in was the game, the sort of senior ex-BR engineers who could see what was wanted were expensive and regarded as unnecessary. Lots of bright, hard-working younger people took their place, but lacked their experience and foresight. NR is no doubt still struggling to recover the required balance.

 

The problem is there is a dire shortage of the required engineers, not just in Network rail but at least in some areas in the contractors that supply them too.  Despite working for a Design House I have been spending a lot of my time over the last year undertaking Project Engineering roles for Network Rail as they don't have enough of their own Engineers to do it. In one recent project the Design house supplied all of the Project Engineers, and of those all apart from me were sub contractors, which I suppose made all the other P.E.'s sub-sub contractors...

 

So the only way Network Rail could bring it back in house would be to nationalise the contractors. And as some of these contractors are multi billion international companies, where the fully integrated railway department is just a part of the huge engineering division, nationalising them just is not an option.

  • Like 2
Link to post
Share on other sites

  • RMweb Premium

The problem is there is a dire shortage of the required engineers, not just in Network rail but at least in some areas in the contractors that supply them too.  Despite working for a Design House I have been spending a lot of my time over the last year undertaking Project Engineering roles for Network Rail as they don't have enough of their own Engineers to do it. In one recent project the Design house supplied all of the Project Engineers, and of those all apart from me were sub contractors, which I suppose made all the other P.E.'s sub-sub contractors...

 

So the only way Network Rail could bring it back in house would be to nationalise the contractors. And as some of these contractors are multi billion international companies, where the fully integrated railway department is just a part of the huge engineering division, nationalising them just is not an option.

 

There is a similar dynamic in other sectors such as defence, energy and maritime, particularly in the more mature economies. Some of the reasons certain power plant construction projects went pear shaped in my time in generation were down to very similar factors (and for those who think it is a UK malaise, the principal contractors were German multi-national companies who were all reliant on Eastern European contract engineers).

Link to post
Share on other sites

  • RMweb Premium

On communication between systems, it isn't necessary for different systems to be able to fully communicate, it is necessary to define the key interfaces and critical information to be exchanged. Industrial control systems do that routinely, most complex industrial processes (many of which are very high hazard such as petrochem, nuclear, defence etc) are designed around distributed control systems controlled by a central control system (a so called system of systems). Each of these distributed control elements may be designed and manufactured by different entities and use proprietary software but by defining the interfaces and comms protocols they all communicate to the degree necessary to operate the wider system. Surely this should be possible for railway rolling stock with standard protocols for various permissives, fault signals and basic controls?

  • Like 2
Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...