2011 Q3 Automatic Train Protection

[BHAR6026]

Hi,
Please find the attached attempted answer to question 3 of module 3(2011).
Please provide review on this.

Thanking you in advance
Madhu

[PJW]

The question was about "accurate braking model"; you seem to start on the wrong foot since your para 1 talks about accurate location of the train. Clearly this information will determine on a particular occasion when it needs to start braking, but it is not directly relevant to the braking model. You say that it has an impact when deciding the braking characteristics- if you meant that the line has changes in gradients and areas of known worse adhesion that then needs to be factored into the braking calculations then you are right, but you totally failed to say that. So perhaps if you placed this para at the end and sufficiently explained that the tolerance of the train knowing exactly where it is so as to feed the correct relevant information into the braking model / how much contingency the braking model needs to cater for the information fed to it being wrong due to positional uncertainty then you may have a valid point, but as it is what you wrote does not seem to be addressing the question.

Para 2. OK but rather waffle than actually giving impression you know about the topic. The essence is that a train may use different braking systems at different speeds during its braking and that the retardation rate therefore varies. Indeed even if just using simple friction brakes, there is the phenomenon of brake fade as the brake blocks warm up and the coefficient of friction reduces. You could have gone further in saying that whereas a train may typically use regen braking for much of its deceleration there can be circumstances when this is not an option and therefore may need to rely on more basic braking methods and therefore not achieve the braking rate that it normally would

Para 3. I the adhesion does not decide what is REQUIRED; it does determine what is ACHIEVABLE

Para 4. OK, but you should have said that the train needs to be able to sense the conditions and therefore the braking model take into account so that it does not assume a more aggressive braking profile than would actually be realisable. Such data as monitoring slip-slide on that vehicle during its journey or obtaining information from the conditions experienced from earlier trains that day when on a specific portion of track could be utilised. Characteristics of the route itself may be quite different (e.g. tunnel section of a metro compared to an open section) and very largely "constant over time" rather than weather related.

Para 5. I consider this really as part of para 2.

Para 6. As per para 1- this is about positional certainty rather than braking model.

Para 7. OK, although curvature not really so relevant as gradient- whereas certainly there is more resistance to motion for a train rounding a curve it certainly depends on its length and I am unaware that we'd ever factor this into the equation used to be certain a train could stop where we need it to.

Para 8. Certainly the speed is a crucial input to the braking model. My understanding of "braking model" is that the stopping place is actually an output from it- given all the input parameters then the outcome is the calculated stopping position. I can conceive though that in some implementations then aiming to get the train to calculate the optimum (e.g. least energy / timed arrival at a specific location) profile rather than the "last sensibly possible application of the brakes", then a braking model could factor in the distance to go available to it but this does need better explanation.

Para 9. OK. Worth separating out factors re how long it take for any braking to commence and the subsequent "build-up" time to achieve the full braking.

Para 10. I don't see this as the braking model per se, but yes this is relevant to the actual performance of the braking calculations and certainly has an effect on resultant headway etc.

So for this first portion of the question you wrote at sensible length and got many of the factors, some not expressed very well and others that look to me as "red herrings" that detracted somewhat. Perhaps 4/7.


PART 2
*1: Not well expressed, better to have said "The effect of a braking model factoring in for guaranteed safety the worst case of various inputs is that the driver will be constrained to drive more cautiously than their knowledge of the actual pertaining conditions would otherwise dictate having the effect of a slower journey time and an adverse effect on headway".


*2. Similar

*3. Agree that this can be the effect of ATP but not that it is the result of the braking model itself. Whereas potentially worth including I would have made that distinction and hence recognise rather tangential to the question actually asked.
I am struggling to find myself what I regard as 10 marks worth of material directly relating to the braking model; I would have expanded my text in *1 to give examples such as-
a) where a driver's route knowledge may be more precise than that fed into the braking model (e.g. a gradient profile that is really falling and then rising but which the train only knows as an average over a long length and possibly in only one of a few increments and therefore always believes more falling / less rising than reality)
and similarly for
b) knowledge of particular train / earlier experience on a previous journey that day / earlier braking experience previously on the same journey etc
etc
then I do think that I would need to expand my interpretation of the question to find enough material. I run the risk of the examiner deciding that I had strayed beyond the set question, so I do think it is worth a sentence of initial justification effectively arguing why consider in scope- "braking model" may well be a phrase that means different things to different things to people in different parts of the industry. However I believe it would have been better to exclude here but place in the last part of the answer as an "operational issue".

*4. Not sure what you are saying here. Seem to imply that ATP will enforce driver's obeying the signal aspects rather than deliberately violating because of their overconfidence that approach release will occur at the last minute. True but not sure it is appropriate here.

HOWEVER
I would certainly have included some text to how the provision of ATP over only part of the route that a driver drives could affect their driving behaviour everywhere; risks of becoming over-dependent on ATP and thus too complacent when driving without it
a) where it is not installed (transition to a different operational mode) or
b) where there is a failure either trackside or on train that requires unplanned driving without it.


Perhaps 3/10 although I admit that I don't see how to get the full 10 marks myself.


PART3.

1. Good (text here is fine given previous discussion on such topic, so if hadn't placed earlier would have needed here instead)

2. Needed more explanation.

3. Good

4. Good

5. Too simplistic. Firstly ATP can certainly be implemented in tunnels but even if transmission by radio / alternative means was not practicable in a particular scenario, this needn't mean there is no ATP. Distinguish between continuous protection and the continuous updating of information. For example ETCS L1 is fundamentally an intermittently updated system with info only being provided to a train as it passes a balise at a signal, perhaps with a length on "infill" from a loop prior; however the train is supervised to profile at all times. It can be restrictive though which relates to your point 1.

6. OK but perhaps slightly more detail

7. Good

8. OK but perhaps also discuss what speed limit may be imposed by the system when ATP inoperative.

I think you should also have discussed
a) migration of ATP geographically during the implementation phase
b) transition at boundaries between ATP fitted areas and non-fitted
c) migration to ATP over time, perhaps with drivers sometimes driving fitted trains and sometimes unfitted trains over the same route or once one route completely migrated then on some days driving trains on other routes

Overall though I think you did the last portion of the question best and would perhaps have got 6/8 for that.

4+3+6 = 13. A pass but only just. Feels about right; f the addition had not made it over the threshold then frankly I'd have gone back to see if I could be more charitable as the overall gut feel (helped by the good ending) was that it was a definite Pass

Hi,
I have given an attempt for 2011 question 5 and I am not sure whether I have placed things properly. Please provide review on this.

Thanking you in advance
Madhu

[BHAR6026]

Thanks for the review PJW. I guess you wrongly posted this review of question 3 2011 (http://irseexam.co.uk/thread-1823.html) in this thread.
Waiting for review of question 5

[PJW]

Sorry hadn't seen there were now two- have moved the placement of this first thread and attached the answer to it. I will try to find time to respond to second but my time very limited at present for a variety of reasons.

Thanks for the review PJW. I guess you wrongly posted this review of question 3 2011 (http://irseexam.co.uk/thread-1823.html) in this thread.
Waiting for review of question 5