(31-05-2010, 05:10 PM)interesting_signal Wrote: I have attempted question 4 (written question) from the 2004 paper. All comments welcome.
For the benefit of anyone reading this who does not have the 2004 paper to hand the question was:
It is intended to introduce a new signal aspect/indication , for use in a degraded mode system, into the 'catalogue' of aspects/indications displayed to the train driver.
Discuss the factors that need to be taken into account when determining the characteristics of such an aspect/indication and the process leading up to its introduction into service.
The key words of the question to me are:
Degraded mode system
Discuss the factors that need to be taken into account when determining the characteristics of
such an aspect/indication and the process leading up to its introduction into service.
Generally I don't like short questions as get less clue what examiner is after; conversely it can give you more flexibility. Back in 2004 they didn't give hints on the mark allocation either so we can only guess.
Your answer clearly gave 5 factors in 1.5 pages; 4 items re process in less than a page; reasonable but I think you should probably have given a bit more on the process.
Characteristics is an important word to explore its possible meaning. I think you did well to think about:
a) the visual impact on the driver,
b) the utilisation to be made of it by the signaller,
c) the interpretation of it by the driver,
d) failure scenario / operational rules,
e) the practicalities from installer and maintainer's viewpoints.
You tuned into the fact that this is a Signalling Principles question rather than Signalling Applications, so we don't want too much on the nuts-and-bolts stuff. You also looked at from different perspectives rather than one element; the question is not one that really lends itself to considering the life-cycle too much but you have alluded to it, so I think that you made a good job of using the scope of the question to advantage.
So we now come onto the actual details of the answer itself
a) Good that you displayed your knowledge odf something similar in a different environment- perhaps disproportionally too much on that though at the expense of other things worth saying. I think I'd have discussed various potential possibilities that might be considered for such an aspect in the UK context. There would be some advantage in not providing a new physical object at all, so perhaps one could discuss whether it would be sensible:
1) to light two existing aspects at once to give a unique indication,
2) to flash the red indication , or even flash the yellow and green alternately?
3) use LED technology to provide another colour from the same aparture used for an existing item- perhaps the top yellow of a 4-aspect signal, the off indication of a GPL or built into a MARI / SARI.
You could have discussed what colours might be possible to utilise, whether the meaning could be conveyed by an illuminated shape rather than just a colour.
Your answer talked about avoiding the risk of potential confusion and that was fine, but perhaps you should also have said something explicitly about the desired range of the indication, the angular spread, day and night visibility etc. You might also have considered whether the aspect should itself, or in conjunction with some other indicator inform the driver which route they are to take; one could argue that in degraded mode the move is so slow that it wouldn't be required, but on the other hand with driver proceeding only as far as the line is clear, it is helpful to let them know which way they are to be looking- particularly at night when there is a maze of pointwork leading to different platforms. Therefore I think that a wider range of titems within your first section would have been beneficial.
I think you were very astute when stating that there needs to be clarity of the precise intention of providing the aspect in the first place; the devil is in the detail. You demonstrated your knowledge of the UK's proposed POSA (I actually thought that it was only points in line of route from post to post that were to be detected, but not 100% sure if that is how it ended up as I don't believe we have ever actually yet commisioned one). You might have expanded this slightly by discussing whether it would prove the signal ahead, whether it would effectively remain off indefinitely or whether there would be a means of at least trying to ensure that it was made "one-shot", whether that be by track circuit, some form of sequential locking with the signals in advance and in rear, by signaller's instruction or so on. One option would be that an aspect would be not much more than a light illuminated by the operation of a switch on the signaller's panel with no frm of interlocking whatsoever; you could have discussed the advantages and disadvantages of something that simple. Even if there is interlocking, how does the signaller operate- would it be by route setting with unique exit button on a panel or icon on a VDU (neither may be easy to fit in an area on a retrofit basis) or would it be from some completely separate system; if so how does the signaller relate the signal(s) being operated in degraded mode with those in the area continuing to operate normally?
The operational rues for the driver are clearly important and inter-related with the above. Perhaps you could have expressed more clearly that a degraded mode aspect has benefits in:
a) improving safety as a result of not being so dependent upon the signaller keying the correct points and the safety critical verbal communication to the driver
b) saving delays.
Requiring the driver to contact the signaller before being given the aspect gives maximum safety but doesn't save delay; using the aspect as an alternative to communicating with signaller reduces delay but has to be a little less safe than if they had also talked, but still better than the current situation of only verbal communication.
Regarding the failure scenario, it was worth saying that it should be designed in a failsafe manner but probably you should have looked wider than that. Would it be worth feeding it from a different fuse and a different cable than the usual signal- probably. From a completely separate location and power supply- almost certainly not. To me it seems obvious and hardly worth saying that if a degraded mode signal failed right-side then the railway would be operated as it would be in the absence of that facility; to me the question is
how independent of the normal signalling it is worth making it to minimise the risk of the same fault affecting this aspect as causes the failure that requires its use at that time.
I think you were right in your last bullet to put a lot of short and sharp questions; may have been better presented by a list of sub bullets after a sentenc of introduction. It is arguably straying a bit outside pure Principles so you were right not to go into detail, but the points were relevant so many and brief was a good way to play it.
======================================================================
The section on process was a bit weaker, but given its length it wasn't bad. I suppose the biggest omissions were:
1) the whole process of getting stakeholder agreement; there are a lot of people to convince all with their individual perspectives. In the UK context I think you should certainly have mentioned the necessary updating of Railway Group Standards (explicitly including the Rule Book of course). There would be the need for risk asseements, HAZOPs etc; in fact a lot would need to have happened before all the good stuff you mentioned.
2) the effect on other things as a result of having introduced. What happens if you take a possession by T2D and disconnect a signal, yet the POSA can still authorise a train to pass the signal. Driver will be looking out for an obstruction on the line but not expecting engineering activities underway- so would it be safe for staff trackside thinking that they had created a green zone? Similarly the signaller's route cards for the installation would need to be updated to acccount for these.
===================================================================
In summary I think that you did this one pretty well. As I said I haven't got my head around awarding marks to individual sections on this question, but it certainly feels like a comfortable Pass and may even be just a Credit, though it is a bit short particularly for the process.
It is definitely considerably better than the one you did from 2009 for the Signet event- if you had come to the Sunday morning sessions on written papers then I'd have claimed the credit for the improvement, but seeing as you didn't then I am denied that!
Seriously it may be worth you looking at
these attachments ; I have now added the 2009 Q1 and Q5 (which I had prepared but decided not to post prior to the event) and these actually cover some similar ground as the first part of this question.
