Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
2008 Q6 INTRO OF NOVEL EQUIPMENT
#5
(26-05-2016, 08:58 AM)dorothy.pipet Wrote: Another attempt for comments please

Part a:
Item 1- The requirement in the UK is generally to demonstrate ALARP; as you have worded the comparison with existing sounds more like the French GAME.  Whilst both are means to an end, there is a difference which can often be significant.  Very relevant to any cross-acceptance argument.

Items 2-4 are all sensible.  The common factor is being able to predict sufficiently accurately with enough confidence.  Novel euipment will always have to be justified by modelling, assumptions etc. etc and are always subject to the "unknown unknown".  Random failures can be estimated pretty well (although it is well known that estimates tend to be unduly pessimistic by a factor of at least 10), but the problem is the unanticipated and therefore may well be a systematic issue that emerges in practice that was not expected at all and might be very significant.  Existing equipment has a "track record" and so there can be far more confidence is that any issues that it might have had would have emerged by now.  This goes quite a long way to explain item 5; it is often not just the "we have always done it like this", but is ecause the individuals have had the bitter experience of "having their fingers burned" from the last thing that was supposed "to be better than sliced bread".  It would have been good for your answer to have brought out this element of uncertainty- it is what "risk" is!
One can attempt to manage what one can anticipate, but there is always could be something suddenly "entering stage left" that just can't be foreseen and so the risk must be considered and accepted for what it is, or the potential product solution rejected (at least until someone else has been prepared to "take the plunge" first).

Your answer should have made clear whether the items listed were from the perspective of the equipment manufacturer or the railway administration or to both; a table with narrow tick boxes in two additional columns would have added little time but been illuminating.  It brings out "who owns the risk" and to some extent this may depend upon the contract.  
[It was interesting to visit Three Bridges depot last week and see the Thameslink Class 700 new trains and learn how the design features of the train were heavily influenced by the need for Siemens to deliver the fleet Availability during the maintenance contract, whereas other facilities that might have been included but relate to TOC rather than Maintenance Company issues were not (no specific requirement in procurement contract, no incentivisation within maintenance contract hence not worth the cost of provision and indeed the additional weight that would be penalised by the energy efficiency incentivisation)].


Part b:

Reasonable and cross-referencing worked ok.

I think you should have included:
  1. Comparison with existing similar products- unlikely that anything is entirely novel, but may be evidence from use within a different industry the remaining challenge is then to know how comparable the environment and that may be as much cultural as environmental
  2. Modelling / Simulation / Analysis
  3. Prototype: accelerated life testing etc
  4. Pilot installation in low risk area (perhaps in sidings adjacent to a mainline so exposed to similar environment but far less impact if failure experienced, low safety risk etc)

Part c:

LED signals potentially use less current, although where backward compatibility is needed then obviously the current draw must be much the same.  Indeed where this is not required, there can be a need for heaters anyway for maintaining the correct colour temperature and prevent being iced over (one of the problems with LED train headlights is that the class covering them needs to include a heater as per car rear windshields!) 

A bit of repetition when comparing advantages of one c/f disadvantages of the other; could have been more efficiently presented.

It is not so much that LEDs are brighter (actually it was quite a challenge to get these to be bright enough- particularly for the yellow since the traditional filter transmits a large proportion of the light, unlike for the Green or Red which are further away from the incandescent light output), it is the angular dispersal that is not so concentrated into the central part of the beam, but more spread out.  Hence "off-axis" LEDs are far, far bighter than conventional and this is actually a disadvantage regarding reading across from parallel roads etc. (but advantage where a single signal is approached around a sharp curve). 

It is not so much that there is not a "bright spot"; more that this has to be explicitly provided in the design of the overall unit (Dorman "eyebrow"or the VMS angled "close viewing segment").
SL35 can be changed faster than an LED module / complete head; the LED MTBF needs to be sufficiently high that the higher MTTR pales into insignificance.  The Howell's "Light Engine" is designed to be a plu-in replacement for an SL35 and so for retrofit is very ttractive as save the need to change the head (and if not careful bring the existing signal structure up to modern standards!) but so far its Reliability has not been anything as good as it should have been and there have been several instances of WSF as well as multiple RSF.  This wouldn't have been known in 2008 when question was set, but would make a very interesting "case study" in the context of this question in 2016- I think they are now up to version 3 the design having been heavily, albeit retrospectively with the benefit of hindsight, criticised.  
In fact within the last year there have been several instances of items which have recently been product accepted to all the most recent standards which have had very significant issues post introduction; smetimes it is not the product per-se but the PADS information did not capture all the necessary "application constraints" - those assumptions that those designing and accepting the product may have made without realising or documenting.  This applies to Frauscher axle counter equipment installed in normal NR location cases, similarly Eltek power supplies and the EbiGate product has had to be withdrawn and the existing commissioned installations taken out of use pending redesign. These demonstrate that as an industry we are not always good at overcoming the challenges involved with the introduction of novel equipment!


As a module 1 question, clearly RISKS and HAZARDS must feature strongly in the answer; I think these with mitigations it would have been worth at least 10 of the last 15 marks and I think it needed more in this section than you provided.

You did identify several items, but it is always wise to word hazards in a way that is specific rather than vague and leave the reader to infer.
A tip to do this, is to word in the form:
IF {something happens / does not happen} THEN {Undesireable Consequence} could happen.

Some other things to have considered:

1. Flashing signal.  If location design not modified then the "filament keep warm" resistors provided for SL35 lamps will prevent an LED signal showing a correct flashing indication thus = leading to potential driver confusion re what aspect is being displayed.

2. Provision of "4 aspect" profile backboard for even a "3 aspect" LED removes a mitigation that was traditionally provided when migrating from 3-aspects into 4-aspects to aid the driver's comprehension of whether the yellow aspect intended as "first caution" or the "inner caution"; could contribute to driver being unduly alarmed by a legitimate yellow or failing to recognise the significance of suddenly encountering a "single yellow".

3. If LED signals intermixed with conventional signals in an unplanned manner, drivers might "read across" lines or "read through" to signals which appear to be brighter than they used to be and ignore the signal which they would be expected to observe and therefore might SPAD.

4. Many years after a wholesale replacement of signals on a route, a common failure mode could affect a large population of them (water ingress due to unexpected failure mode such as cracking after thermal cycling, electro-chemical deterioration of electronics) unexpectedly.  If not remaining alert for this possibility and be monitoring for early signs, there could be a sudden need to replace a large number for which there is no financial provision, possession availability, replacement equipment available. 

5. If a traditional GPL which had a backlight is replaced by an LED version, in some circumstances (at night without the area being significantly lit), the train driver may not be able to see the signal behind which they are due to reverse.  Similarly a local signaller or shunter may also have been reliant upon identifying the position of the train relative to that signal and would no longer be able to do so.
PJW
Reply


Messages In This Thread
2008 Q6 INTRO OF NOVEL EQUIPMENT - by losler - 16-08-2010, 12:56 AM
RE: 2008 Q6 INTRO OF NOVEL EQUIPMENT - by PJW - 16-08-2010, 06:28 AM
RE: 2008 Q6 INTRO OF NOVEL EQUIPMENT - by PJW - 18-08-2010, 09:47 PM
RE: 2008 Q6 INTRO OF NOVEL EQUIPMENT - by PJW - 30-05-2016, 11:37 AM

Forum Jump:


Users browsing this thread: 1 Guest(s)