Too much variation in new-build rolling stock?

[D365]

Hitachis class 8xx are the most expensive trains in the UK but that didn't stop them being revealed with fundamental problems thats going to take years to fully resolve.
The core issue here is we should have just settled on one design got it right and mass produced them rather than letting each operator do its own thing specifying what it wants.

Your own example hinders your argument - how could one design have satisfied the needs of the Hitachi 8xx and CAF 19x orders? They were built to serve very different markets.

It’s a nice thought, but surely one that isn’t ever going to be feasible. Look at how many different Class 19x, 7xx, 8xx variants are being turned out. Too many small quantities and incremental tweaking.

 

[Snow1964]

This is too much re-engineering the basic design, causing small production runs.

Should have been more universal designs, where only the livery and seat configuration and upholstery colour were the variables.

Ultimately only really needed one basic design for high speed, a regional (secondary main line), an outer suburban electric unit, an inner suburban / local train. Different lengths by adding more intermediate cars to a standard design (mainly seats, one small toilet)

 

[D365]

As an engineer I find that a lot of energy goes into the human factors side, due to the level of involvement with drivers/guards. Consequently this is one of the areas where rolling stock introductions, more often than not, get delayed.

 

[popeter45]

One suggestion at minimum is bundling all the variants into sub classes rather than full new classes

Is their really enough different between a 800, 801 and 802 to not class then as e.g. 800/0, 800/1, 800/2 etc?

 

[HamworthyGoods]

One suggestion at minimum is bundling all the variants into sub classes rather than full new classes

Is their really enough different between a 800, 801 and 802 to not class then as e.g. 800/0, 800/1, 800/2 etc?

You could argue the same with 314s and 315s surely or 317s and 318s etc

 

[Irascible]

Most of the problems seem to be in the abysmal state of TMS software compatibility - there's no excuse for the state it's in other than shoddy management.

 

[popeter45]

also in general a unified coupler system and multible working system would go a long way

no more need for complex translator vans or having to swap out control systems if a DVT is planned to use a diffrent type of locomotive etc

for the multible working a software defined ethernet based system would prob be the best solution

 

[Irascible]

also in general a unified coupler system and multible working system would go a long way

no more need for complex translator vans or having to swap out control systems if a DVT is planned to use a diffrent type of locomotive etc

for the multible working a software defined ethernet based system would prob be the best solution

Maybe we should be pushing for a standards body to regulate this sort of thing!

( I have a RSSB account, before anyone takes that at face value )

--

TMS "coupling" could be as easily defined as a physical coupling if there was a desire to do it. Software standards do not seem to be the best, worryingly.

 

[MattRat]

TMS "coupling" could be as easily defined as a physical coupling if there was a desire to do it. Software standards do not seem to be the best, worryingly.

That's probably an understatement. I swear a few years back the NHS computers got hacked and it came to light they were still using Windows Vista or something like that. I dread to think what the railways are using, probably something that can be traced back to the Commodore 64.....

 

[RobShipway]

Maybe we should be pushing for a standards body to regulate this sort of thing!

( I have a RSSB account, before anyone takes that at face value )

--

TMS "coupling" could be as easily defined as a physical coupling if there was a desire to do it. Software standards do not seem to be the best, worryingly.

Part of the problem and I say this as someone that has worked in both IT Software and Hardware for 30 years, is the systems that the software for the trains are running on.

Not sure if it is still the case, but the class 220/221 Voyagers certainly when introduced where running on Windows 98, even though by the time of their introduction Windows 2000 had been around a few years!

Not sure what any of the newer trains are running on, but I believe some are running on Linux systems, which if it is the case should give the units better ability to adapt and have software on them that can allow multiple working between different units. The only situation then, is the physical coupling when has been discussed many times in many threads in these forums.

 

[ComUtoR]

Should have been more universal designs, where only the livery and seat configuration and upholstery colour were the variables.

Ultimately only really needed one basic design for high speed, a regional (secondary main line), an outer suburban electric unit, an inner suburban / local train. Different lengths by adding more intermediate cars to a standard design (mainly seats, one small toilet)

I think you're forgetting the important part of train design. The carriages are almost meaningless. It's the pointy end that needs constant updating.

 

[Turtle]

I think you're forgetting the important part of train design. The carriages are almost meaningless. It's the pointy end that needs constant updating.

Ah, the dear old Class 460 with a real pointed nose.

 

[D365]

I think you're forgetting the important part of train design. The carriages are almost meaningless. It's the pointy end that needs constant updating.

Indeed - a train cab is the interface between driver and systems.

 

[Irascible]

Part of the problem and I say this as someone that has worked in both IT Software and Hardware for 30 years, is the systems that the software for the trains are running on.

Not sure if it is still the case, but the class 220/221 Voyagers certainly when introduced where running on Windows 98, even though by the time of their introduction Windows 2000 had been around a few years!

Not sure what any of the newer trains are running on, but I believe some are running on Linux systems, which if it is the case should give the units better ability to adapt and have software on them that can allow multiple working between different units. The only situation then, is the physical coupling when has been discussed many times in many threads in these forums.

It doesn't really matter what they run, it's only how they talk to everything else that matters - prime example of how to actually do it, the work of the IETF. I don't really understand why the RSSB hasn't jumped on this already, but I guess it's just been a matter of priorities & unfortunate timing. The Voyagers would have fallen in a hole in the system no matter what given when they were built, but there's little excuse for later incompatibilities.

 

[D365]

Part of the problem and I say this as someone that has worked in both IT Software and Hardware for 30 years, is the systems that the software for the trains are running on.

Not sure if it is still the case, but the class 220/221 Voyagers certainly when introduced where running on Windows 98, even though by the time of their introduction Windows 2000 had been around a few years!

Not sure what any of the newer trains are running on, but I believe some are running on Linux systems, which if it is the case should give the units better ability to adapt and have software on them that can allow multiple working between different units. The only situation then, is the physical coupling when has been discussed many times in many threads in these forums.

As far as I can tell, the Voyagers used Windows CE as the train management operating system. Windows 98 and Windows 2000 were still very new at the time, Bombardier wouldn't have had the time to develop SIL (Safety Integrity Level) compliant software.

Besides, it's not just the software, but the fundamental electric architecture of each manufacturer that precludes interoperability.

== Doublepost prevention - post automatically merged: 25 Dec 2022 ==

It doesn't really matter what they run, it's only how they talk to everything else that matters - prime example of how to actually do it, the work of the IETF.

Please can you define IETF.

 

[popeter45]

Please can you define IETF.

Internet engineering task force essentially the standards body for the internet

 

[Andrew Holland]

That's probably an understatement. I swear a few years back the NHS computers got hacked and it came to light they were still using Windows Vista or something like that. I dread to think what the railways are using, probably something that can be traced back to the Commodore 64.....

It was Windows XP not Vista, so even older as XP dates back to 2001, Vista was 2006.

 

[Irascible]

Please can you define IETF.

Yes, I should have - as was mentioned, the Internet Engineering Task Force is a standards organisation that draws on & heavily involves software & hardware manufacturers in the setting up & developing of the standards that make the Internet work. It's not even really a body, you just have to demonstrate technical competence to contribute.

The RSSB also involves industry - which makes perfect sense - but for whatever reason there's apparently been no guidance in this area.

 

[david1212]

This is too much re-engineering the basic design, causing small production runs.

Should have been more universal designs, where only the livery and seat configuration and upholstery colour were the variables.

Ultimately only really needed one basic design for high speed, a regional (secondary main line), an outer suburban electric unit, an inner suburban / local train. Different lengths by adding more intermediate cars to a standard design (mainly seats, one small toilet)

Exactly.
Common drivetrain, common core body and components so minimising the variety of parts required and type training for both maintenance and operation. Overtime the stock can be moved between operators where if already competent on the type no further training required.

As a specific fleet that has been shuffled around the 150's I guess are the last that filled this criteria.
Further though the bodyshell was also used for the 455 third rail EMU and 318 overhead EMU.
All followed on from the 317 although under the skin I am unsure of differences.

Part of the problem and I say this as someone that has worked in both IT Software and Hardware for 30 years, is the systems that the software for the trains are running on.

Not in anyway railway related but I am expected still to keep NT4 systems running. I can not recall dealing with a DOS system this year but did last year.

Not sure if it is still the case, but the class 220/221 Voyagers certainly when introduced where running on Windows 98, even though by the time of their introduction Windows 2000 had been around a few years!

Not sure what any of the newer trains are running on, but I believe some are running on Linux systems, which if it is the case should give the units better ability to adapt and have software on them that can allow multiple working between different units. The only situation then, is the physical coupling when has been discussed many times in many threads in these forums.

Whatever the platform the time period from initial concept to fleet withdrawal should be presumed as at least 40 years. Expecting still to be using the same components and software platform is not realistic unless a lifetime stock of modules are manufactured together with the initial production. ( Back to my day job 10+ years ago a system was installed with 30 years expected life. Within this are 12 PC's with common base hardware then different add-in cards. Just one spare of each was purchased ..... the recommendaton was 36 motherboards, software loaded drives & power supplies plus sets of add-in cards ) The logical solution is a modular design where ' black boxes ' can be replaced with a new design but the same functionality or an enhanced functionality without full system replacement.

 

[miklcct]

Not in anyway railway related but I am expected still to keep NT4 systems running. I can not recall dealing with a DOS system this year but did last year.

This is the problem of using proprietary software which can fall out of vendor support. If only open source software is used, even if it is out of vendor support one can still source third party support if needed.

 

[D365]

If only open source software is used, even if it is out of vendor support one can still source third party support if needed.

Very difficult to introduce open source software in a railway environment… somebody has to be responsible for the safety aspects. That doesn’t come free.

 

[306024]

Look at all the different disability ramps held at stations for all the different class of trains. I’ve seen pictures at Ely and Stirling of a row of ramps, wonder which station holds the record.

 

[david1212]

This is the problem of using proprietary software which can fall out of vendor support. If only open source software is used, even if it is out of vendor support one can still source third party support if needed.

The primary issue I face is rustling up hardware compatible in this example with both NT4 and the add-in hardware rather than the base software platform having been released and the custom software written ~25 years ago.
The software could, maybe in stages, relatively easily and quickly be rebuilt to run on current hardware and Windows 10 ( The custom software platform does not ( yet ) support Windows 11 ). A bigger issue is sourcing add-in hardware with both the same functionality and easy integration with the rebuilt software. Inevitably either rewiring or a conversion box would be required.
The biggest issue for the customer though is both the downtime and cost to fully revalidate the system to prove the system functionality is unchanged including fault conditions.

 

[Irascible]

Very difficult to introduce open source software in a railway environment… somebody has to be responsible for the safety aspects. That doesn’t come free.

While there are safety-criticical versions of ( orignally ) OSS, I agree - but open standards are the more important thing than the software itself. It doesn't matter what the software is as long as it can talk to it's peers. Open standards sitll enable the beloved market competition.

 

[popeter45]

How I would best correlate open standards in a railway friendly term would be for example loading and track gauge

The W6 loading guage and standard gauge track are open standards everyone can use with their closed source designs so everything on the railway works together

 

[Irascible]

How I would best correlate open standards in a railway friendly term would be for example loading and track gauge

The W6 loading guage and standard gauge track are open standards everyone can use with their closed source designs so everything on the railway works together

The most obvious ones are couplers & braking systems. Imagine if similar fragmentation had happened there as we get with TMS & so on these days - we're seeing a bit of it now, but imagnie every manufacturer of rolling stock in the railway heyday fitting their own devices.