Electrical Obsolescence on rolling stock, getting worse or better?

[Servotech Rail]

In my role at Servotech I'm always amazed at the sheer amount of electrical obsolescence issues we see on a weekly basis from our rail customers.
There's everything from CCTV, HVAC Plc's, DCU's, PIDs all the way down to old relay boards and from what I've seen from the new rolling stock (Innotrans14) they have even more electronics than ever before so it looks like things won't be getting any better as time goes by.

I'm trying to put together an "Obsolescence database" so that I can potentially help more people so I'd love to hear what issues are out there and what fleet\class of vehicle it relates to.

 

[Clip]

Are you saying these should be replaced more frequently? If so how are you proposing to continually pay for these upgrades regularly?

 

[D365]

One of the main problems that I've come across are onboard PA systems which have failed display screens or are simply not useful at all.

 

[Servotech Rail]

One of the main problems that I've come across are onboard PA systems which have failed display screens or are simply not useful at all.

Displays either from PIS\PAS systems or driver display panels tend by design to by more fragile than some of the other parts we see, also the technology in displays just like TV's etc becomes obsolete faster than than other parts, still a lot of it can still be overhauled rather than replaced.

 

[theageofthetra]

PIS faults are the most common faults I see on the traction I sign. Other more serious faults of course means the unit doesn't enter service so I guess that may be partly to blame.

 

[Bodiddly]

Electrical obsolescence in the railway industry tends to be drawn out just that little bit longer as cost is much more of an issue due to franchising. If you take signal heads for an example. Colour light signaling schemes pre LED were designed to last 25 years +. As new LED signaling is used, that should come down slightly as I cannot see these new signal heads functioning for 25 years without some major component going kaput. I maybe wrong but on board PIS which utilizes relatively old dot matrix technology isn't about to disappear for some time. I would have thought the new train builders would be using more LCD or LED screens for extra advertising income.

 

[Servotech Rail]

PIS faults are the most common faults I see on the traction I sign. Other more serious faults of course means the unit doesn't enter service so I guess that may be partly to blame.

Yes I would agree that faults on PIS equipment is one of the major fault issues these days, but cctv problems with recorders, controllers and cameras would be right up there as well.
--- old post above --- --- new post below ---

Electrical obsolescence in the railway industry tends to be drawn out just that little bit longer as cost is much more of an issue due to franchising. If you take signal heads for an example. Colour light signaling schemes pre LED were designed to last 25 years +. As new LED signaling is used, that should come down slightly as I cannot see these new signal heads functioning for 25 years without some major component going kaput. I maybe wrong but on board PIS which utilizes relatively old dot matrix technology isn't about to disappear for some time. I would have thought the new train builders would be using more LCD or LED screens for extra advertising income.

LED's are quite good not they are not indestructible but I think the MTBF will be better on these new parts.

 

[SpacePhoenix]

Is much of the electronics on modern stock design "plug & play" style so that an electronics panel/board can be unplugged and swapped with an updated panel/board?

 

[rebmcr]

I never understood why the rail industry doesn't standardise on COTS (commercial off-the-shelf) technology.

 

[Servotech Rail]

I never understood why the rail industry doesn't standardise on COTS (commercial off-the-shelf) technology.

In principal it's a good idea, however with so many oem's offering their own designs it would be unworkable.

 

[ComUtoR]

Other more serious faults of course means the unit doesn't enter service so I guess that may be partly to blame.

I've mentioned this before. The safety critical systems are much more robust and meet higher requirements and receive regular maintenance.

I think the other problem is procurement. The PIS has been broken from day one but the sheer cost to replace the entire system and that we are now contracted for the long term means its almost impossible to replace the crap we get given.

 

[Servotech Rail]

I've mentioned this before. The safety critical systems are much more robust and meet higher requirements and receive regular maintenance.

I think the other problem is procurement. The PIS has been broken from day one but the sheer cost to replace the entire system and that we are now contracted for the long term means its almost impossible to replace the crap we get given.

Passenger information systems can be quite complex and are integrated right through the train\tram. This one is our BTROS Sacu\Kacu test rigs.
The actual design\placement of the unit in this case caused water ingress in to the SACU's.

 

[ComUtoR]

Er.. thanks for the pretty picture but that still didn't really address my point at all.

You can test the systems as much as you wish but you need to test it with a live installation to see the issues in real terms.

If its been designed and placed poorly its too late when its been stuck into the trains. By then its too late. Then we have the situation we have in that nothing is gonna change. It will remain broken.

 

[Servotech Rail]

Er.. thanks for the pretty picture but that still didn't really address my point at all.

You can test the systems as much as you wish but you need to test it with a live installation to see the issues in real terms.

If its been designed and placed poorly its too late when its been stuck into the trains. By then its too late. Then we have the situation we have in that nothing is gonna change. It will remain broken.

I don't agree with the "nothing is gonna change" statement.
Pretty picture to one side, what we have done is highlight the water ingress to the TOC and they have identified the source of the ingress and put in measures to protect the sacu unit and while still not perfect we are seeing less units being returned for repair.

 

[AM9]

Non-safety critical electronics on trains still seems relatively immature. There's nothng particularly difficult in designing equipment such as PIS to be reliable, maintainable and testable. Whilst there are aspects of the environment that demand certain a robustness of equipment, none of those are that difficult to achieve with current technology, and within a cost-sensitive market.
The rail environment within the passenger space is benign in temperature and mechanical stress. The major mechanical design objectives are concerning protection against tampering, theft and malicious damage. The electrical environment is quite agressive but easily dealt with using good EMC design practice.
Maintainability is clearly a critical requirement in terms of overall stock availabilty. The nature of electronic equipment within large capital transport vehicles is that repairs would normally be performed away from the operational level. This requires a level of provisioning of replaceable items available at depot level where system faults can be rectified within normal (overnight) servicing schedules.
An essential element of the reliability and maintainability regime is testability. That means the ability to quickly determine whether an item is functional in situ as part of the replacement process. Self-diagnosis is preferable and is dependent on appropriate functional partitioning in design aswell as effective in-situ self-test routines.
The above support strategy requires a system design that has parallels with the computer industry. Particularly, system interfaces should be industry standards as far as possible. Typically, Ethernet is capable of delivering a universal interface between most non-safety critical items on board with appropriate ruggedisation. Similarly, distributed power systems assist EMC performance for the more sensitive items, yet using modern electronic designs, have a low incremental cost.
In a more general industry sense, none of the above is new. It has been good practice in the aircraft industry for decades, well before the advent of low-cost defacto interfaces created by the information technology industry made it an imperative.

 

[chubs]

I would have thought the new train builders would be using more LCD or LED screens for extra advertising income.

The Dutch (NS anyway) already decommissioned the old LED dot matrix screens on their trains (which they had long before us) and retrofitted LCD screens that scroll the list of stops of the train, expected arrival at each station, minutes delay, connections etc.

We'll probably get it at much greater expense when it has become obsolete.

 

[Nym]

To answer the title,

Worse.

More complex interfaces.
IP issues.
Higher deviations from standards when used in their infancy.
RIA standards still apply to most pre 90s stock, even if TOCs don't think so.
Worst of all, 90% od electronics manufacturers don't understand trains or railways.
And a much longer list...

 

[AM9]

To answer the title,

Worse.

More complex interfaces.
IP issues.
Higher deviations from standards when used in their infancy.
RIA standards still apply to most pre 90s stock, even if TOCs don't think so.
Worst of all, 90% od electronics manufacturers don't understand trains or railways.
And a much longer list...

As a lay person, what do the RIA standards cover please?

 

[Servotech Rail]

As a lay person, what do the RIA standards cover please?

Railway industry standard

 

[AM9]

Railway industry standard

Thanks. What is their scope then?

 

[Servotech Rail]

Thanks. What is their scope then?

I'm sorry, I don't know the answer to that.

 

[Nym]

As a lay person, what do the RIA standards cover please?

They're a series of "Railway Industry Association" recommendations and standards that have now been superseded by several British Standards and Euro Norms.

For example,
RIA-12 - General specification for protection of traction and rolling stock electronic equipment from transients and surges in DC control systems.
Has been replaced by EN50155 - Railway applications. Electronic equipment used on rolling stock

However, EN50155 is nowhere near as stringent as RIA-12, due to the evolution of rolling stock moving from MA powered auxiliaries to static converter powered auxiliaries, the surge and transient performance especially has been rather watered down.

Other recommendations, such as RIA-10 have been partially replaced by various ENs, but the RADOX / Mutlicore cable "Standard" forms are now not part of any other standard, so internal company standards have picked up to continue standardisation. Where mainline have now pretty much entirely moved to conduited jumpers or using a single source cable, from a supplier such as Huber+Shruner, where previously a RIA-10 specified BRB or LUL part number will have been specified for multicore jumpers to a standard design, available from several companies, such as AEI and Pirelli.

A list of some of the RIA specifications is available here http://www.riagb.org.uk/ria-technical-specifications/

 

[AngusH]

(Deleted, this comment added nothing, sorry)

 

[AM9]

They're a series of "Railway Industry Association" recommendations and standards that have now been superseded by several British Standards and Euro Norms.

For example,
RIA-12 - General specification for protection of traction and rolling stock electronic equipment from transients and surges in DC control systems.
Has been replaced by EN50155 - Railway applications. Electronic equipment used on rolling stock

However, EN50155 is nowhere near as stringent as RIA-12, due to the evolution of rolling stock moving from MA powered auxiliaries to static converter powered auxiliaries, the surge and transient performance especially has been rather watered down.

Other recommendations, such as RIA-10 have been partially replaced by various ENs, but the RADOX / Mutlicore cable "Standard" forms are now not part of any other standard, so internal company standards have picked up to continue standardisation. Where mainline have now pretty much entirely moved to conduited jumpers or using a single source cable, from a supplier such as Huber+Shruner, where previously a RIA-10 specified BRB or LUL part number will have been specified for multicore jumpers to a standard design, available from several companies, such as AEI and Pirelli.

A list of some of the RIA specifications is available here http://www.riagb.org.uk/ria-technical-specifications/

Thans for the link. It seems that most of them involve the CE & RE of traction equipment and auxiliary services. As you say, the modern trends of solid state inverters etc., are following more conventional (i.e. non-rail) EMC requirements. In reality, the task of designing electronic line replaceable items (e.g. displays as was originally discussed) is much easier today with common auxiliary power supply buses and well protected DC-DC conversion within an EMC compliant enclosure.

 

[Nym]

Well, Euro Norms of railway power standards...
That are a lot easier to comply with.

 

[hedpe]

Class 313 trains use a lot of contactors and relays which of course have mechanical moving parts and these can deteriorate in performance over time. They are regularly checked on exams. The class 313 was designed and built in the 70's and there were no micro processor systems built in. The wheel slip/slide system is a large rack of circuit boards that reads frequencies and (i believe) uses gates to determine if a wheel is slipping or sliding. New trains have fewer mechanical components and more solid state hardware. Previously equipment would "talk" to each other by energizing or de-energizing a specific wire at 110v, now it's very low voltage communication (Ethernet) or fibre optic even. Low voltage wires are easier to deal with. Solid state equipment doesn't gradually deteriorate and should theoretically last longer.

These trains now have cab air cooling, GSMR, passenger counting, energy metering and data recorders fitted. These components were all added gradually relying on the space and power already available from the train. Now they are not totally unreliable but things such as accessibility is compromised. Newer trains are a lot more modular whereas retrofitted equipment to older trains has bits and pieces hidden behind hard to reach panels or seating.

As trains get older, spares can become increasingly hard to get hold of. Third parties (ie UNIPART) will manufacture components, but you have to pay the price.

Then there can be other reasons ie contractors

 

[notlob.divad]

Don't work in rail, but do work in a Electronics / Electrical / Computing in heavy industry and a couple of my ex colleagues have moved across into rail. So there are similarities.

Part of my job is the "cannibalization" of working components from systems so we have spares to ship halfway around the world to keep other process going the last few years until their major overhauls. Another part is the dealing with the mad panic when places overlook their obsolescence issues and immediately demand a solution which may not be forthcoming.

It is interesting that in the railway industry you are saying it is the PIS/CCTV etc that is the issue first, and I would agree that this is my experience. The problem of obsolescence is an ongoing battle and in general seems to get worse and worse. Whilst many of the newer components may have a longer MTBF, when they do fail it is usually harder to get a direct replacement. When a relay goes, if you cannot get the same relay, you can usually whip it out and replace with a more modern but similar functioning component. However when a microprocessor goes, you pretty much have to get the same one otherwise the software doesn't work.

In an ideal world you would rip everything out and put brand new in at the first available opportunity, to safeguard you for the next xx years. However it is just not practical and the cost spirals, but at some point you just have to bite the bullet and go for a major upgrade otherwise you end up wasting more money in the long run. At least with the number of vehicles on the rails if you do a major upgrade the upfront design and testing costs are worthwhile. We sometimes end up doing a major upgrade that only gets used once before it is itself obsolete.

It is a thankless task and my thoughts go out to anyone who has to deal with these issues on a regular basis.

 

[Servotech Rail]

Class 313 trains use a lot of contactors and relays which of course have mechanical moving parts and these can deteriorate in performance over time. They are regularly checked on exams. The class 313 was designed and built in the 70's and there were no micro processor systems built in. The wheel slip/slide system is a large rack of circuit boards that reads frequencies and (i believe) uses gates to determine if a wheel is slipping or sliding. New trains have fewer mechanical components and more solid state hardware. Previously equipment would "talk" to each other by energizing or de-energizing a specific wire at 110v, now it's very low voltage communication (Ethernet) or fibre optic even. Low voltage wires are easier to deal with. Solid state equipment doesn't gradually deteriorate and should theoretically last longer.

These trains now have cab air cooling, GSMR, passenger counting, energy metering and data recorders fitted. These components were all added gradually relying on the space and power already available from the train. Now they are not totally unreliable but things such as accessibility is compromised. Newer trains are a lot more modular whereas retrofitted equipment to older trains has bits and pieces hidden behind hard to reach panels or seating.

As trains get older, spares can become increasingly hard to get hold of. Third parties (ie UNIPART) will manufacture components, but you have to pay the price.

Then there can be other reasons ie contractors

We've seen a few wheel slip slide systems coming to us for repair over the years, anything from 40 year old relay Mil\Rail spec. based boards to the slightly more update to date WSP racks as shown in our test rig below.

 

[randyrippley]

What all this points to is the stupidity of the current system which results in small orders of bespoke rolling stock for individual TOCs or routes. How many different classes and subclasses of coaching stock now run on our railways, often with discrete components? What you end up with is small volumes of unique parts which no-one can be bothered to make.
Proper planning would have resulted in large orders of perhaps five or six designs of stock, each with a standard configuration allowing decent component production runs, and ongoing availability of those components

 

[Servotech Rail]

What all this points to is the stupidity of the current system which results in small orders of bespoke rolling stock for individual TOCs or routes. How many different classes and subclasses of coaching stock now run on our railways, often with discrete components? What you end up with is small volumes of unique parts which no-one can be bothered to make.
Proper planning would have resulted in large orders of perhaps five or six designs of stock, each with a standard configuration allowing decent component production runs, and ongoing availability of those components

In the long run this idea is unworkable, the oem's need to development new components and get companies to buy the new parts, long production runs and component with long availability would do little for technological innovation.