"Type training" ops manual changes

[Bob Llewellyn]

You're probably right, I just see problems with the stats you are using. Adjusting the assumptions until the results are within expected bounds isn't right.I would be more interested to see (and much more convinced by) stats that showed RAA accidents vs. GA private flying accidents, for the same years. Even then you would have doubt about the effect of RAA vs. PPL training, unless you had the figures for how many of the RAA pilots had a PPL and could adjust for that.

Very good, but I start by considering the possible - NOT the expected bounds.

 

 

[Bob Llewellyn]

I must admit,I'd love to get my hands on some drawing and specs for Drifters.I see more clearly what you are saying now, and I agree some maintenance training can't hurt, but I suspect it will improve things very little.

The AUF concentrated on pilot training, and the accident rate plummeted. Now the accident rate is rising. If it is not hardware-related, how did pilot training screw up in 2009 and subsequently? Is the Human Factors training a killer?

 

 

[aro]

Pre flight inspections are a maintenance procedure - by your argument, eliminating pre-flights will increase safety. Well, off you go then!

I'm not arguing for eliminating maintenance. Some maintenance is obviously necessary. What I am suggesting is that if you design so less maintenance is required, safety and reliability are improved. Part of that is because whenever maintenance occurs, errors are possible. Most people who have been involved in aircraft maintenance could give examples of problems that occurred after maintenance - but to repeat, I am not arguing that that means the maintenance should not have occurred.

 

 

[Bob Llewellyn]

I'm not arguing for eliminating maintenance. Some maintenance is obviously necessary. What I am suggesting is that if you design so less maintenance is required, safety and reliability are improved. Part of that is because whenever maintenance occurs, errors are possible. Most people who have been involved in aircraft maintenance could give examples of problems that occurred after maintenance - but to repeat, I am not arguing that that means the maintenance should not have occurred.

No aeroplane can be designed for infinite life, as no materials give the necessary strength to weight ratios within their fatigue (metals / plastics) or creep (timber) limits. Therefore, the most structurally efficient aircraft will fail "all at once and nothing first..."; to prevent a nasty surprise, surely this means that one should inspect everywhere?

If one completely eliminates fasteners of any sort - which means some sort of FRP, including the engine mount - and creates the nanobots necessary to detect the onset of post-impact fatigue damage (see FAA, rudder doublets) - and puts up with the FRP undercarriage springs regularly failing due to the nature of the material - you still have fuel pumps, fuel hoses, fuel fittings, an engine, a propellor, fasteners... but that will be a minimum-inspection aeroplane. A composite glider would be the go... but how do you take off?

 

 

[aro]

Very good, but I start by considering the possible - NOT the expected bounds.

Your RAA hardware failure rate is based on the assumption that the RAA human factors accident rate is 3x that of GA. Where do you get that figure? You started with equivalent rates, which gave 96% of accidents caused by hardware failure. Then you seemed to conclude that 3x gave a reasonable figure. Where does 3x come from, if not from the result?

 

If the GA figures include all commercial GA flying, under an AOC and overseen by company SOPs etc. I would be surprised if the RAA human factors accident rate was only 3x that of GA. Since the figure translates directly into the result, you can't just pick a figure and assume it.

 

 

[aro]

No aeroplane can be designed for infinite life, as no materials give the necessary strength to weight ratios within their fatigue (metals / plastics) or creep (timber) limits. Therefore, the most structurally efficient aircraft will fail "all at once and nothing first..."; to prevent a nasty surprise, surely this means that one should inspect everywhere?

Which aircraft allow you to inspect everywhere? All aircraft have some areas that can't reasonably be inspected.

 

But what I said was: "Some maintenance is obviously necessary. What I am suggesting is that if you design so less maintenance is required, safety and reliability are improved."

 

Surely you would have to agree with this? Don't engines with thousands of hours TBO instead of hundreds increase reliability?

 

 

[M61A1]

The AUF concentrated on pilot training, and the accident rate plummeted. Now the accident rate is rising. If it is not hardware-related, how did pilot training screw up in 2009 and subsequently? Is the Human Factors training a killer?

The way it's applied in this country.................... could be?

There are always blips and dips in stats that hide trends, or make non-existent ones.

 

I should also mention that when I say some training could help.......I mean as long as it's not an excessive burden, in which case we'll all be safe, because the only way to afford to aviate will be on a 737 with a red rat on the tail.

 

 

[Bob Llewellyn]

Your RAA hardware failure rate is based on the assumption that the RAA human factors accident rate is 3x that of GA. Where do you get that figure? You started with equivalent rates, which gave 96% of accidents caused by hardware failure. Then you seemed to conclude that 3x gave a reasonable figure. Where does 3x come from, if not from the result?If the GA figures include all commercial GA flying, under an AOC and overseen by company SOPs etc. I would be surprised if the RAA human factors accident rate was only 3x that of GA. Since the figure translates directly into the result, you can't just pick a figure and assume it.

The RAAus training syllabus is accepted by CASA; and 20 hours minimum for day VFR to the equivalent of a Restricted PPL (which, pre competency-based days, could be had in 35 hours) suggests that it should be of similar rigour. If you have data otherwise, please notify the Ops Manager and CASA. We also have BFRs, just like GA.

 

Now, if you read the document, I set one boundary at RAA pilots having no less an accident rate than GA; which, you state above, you consider excessively low. I then point out that the residuum of RAAus fatal accidents, some 96% of them, would be an extraordinarily high level to set at the door of hardware failure. I did not do the simple maths in the document, but your calculator will no doubt tell you that if the RAAus hardware failure rate is the same as GA, then the human factors RAAus accident rate must be 14 times the GA rate; and the RAAus training regime, ~1/14th as effective. Hands up anyone who believes that the RAAus pilots (a lot of whom are also GA) are this woeful?

 

Halfway between the boundaries of the same HF accident rate as GA, and the same hardware accident rate as GA, lies a HF accident rate of 7 times that of GA. As a lot of RAAus pilots also have had GA training, and based on those I have met, I do not believe that the non-GA trained RAAus pilots are vastly less competent than the GA trained ones, within the operational limitations of 95:55. A similar situation exists in the Gliding Federation of Australia.

 

As the RAAus training syllabus is CASA acepted, AND the RAAus accident rate spent several years below the GA accident rate at a similar state of maturity, AND the GFA have over 50 years of satisfactory HF accident rate despite putting most pilots through in less than 20 hours, AND a lot of RAAus pilots also have GA training, I consider that assuming three times the GA HF accident rate is unreasonably conservative. I do not see any reason supporting your arguments to the contrary.

 

If you have anything more than a baseless opinion on which to condemn RAAus pilots, I reiterate that you should communicate it to the ops manager and CASA immediately; and also present the justification for your apparently baseless insistence that RAAus experimental homebuilts - and aircraft built to partial design standards, such as the Skyfox - are just as safe as Cessna 172s maintained in accordance with GA practise.

 

My ANALYSIS of the DATA is now fully on the table. Where is yours?

 

 

[Bob Llewellyn]

Which aircraft allow you to inspect everywhere? All aircraft have some areas that can't reasonably be inspected.But what I said was: "Some maintenance is obviously necessary. What I am suggesting is that if you design so less maintenance is required, safety and reliability are improved."

 

Surely you would have to agree with this? Don't engines with thousands of hours TBO instead of hundreds increase reliability?

NO - study MIL-HDBK-217; "Probability and Statistics for Engineering and the Sciences"; failure modes and effects analysis.

Also, the 2,000 hr TBO Lycoming has been maintained at least 80 times in that period, with the cylinder heads reconditioned at 1,000 hrs. By your "reduce maintenance" argument, a Rotax 503 run 300 hours, with only 12 maintenance exercises in that period (and no head jobs!) must be safer.

 

 

[Bob Llewellyn]

The way it's applied in this country.................... could be?There are always blips and dips in stats that hide trends, or make non-existent ones.

I should also mention that when I say some training could help.......I mean as long as it's not an excessive burden, in which case we'll all be safe, because the only way to afford to aviate will be on a 737 with a red rat on the tail.

Yeah, so I took 9 year's worth of data. Not much scatter, either.

I think a correspondence course for less than $200- total, would suffice for the L2s interested. The issue is presenting the necessary engineering information in a way that you don't have to be an engineer to use (but you will have to be able to read a chart / graph:cheers:)

 

Note: i can say no more on this head, without violating the non-commercial site rules...

 

 

[M61A1]

Just had a look at the ATSB reported ultralight incidents for 2009,not a lot to go on,

 

- 33 reported (I know that's not all of them)

 

- 6 fatalities including one double.

 

- All 6 fatalities recorded as "Operational- collision with the ground"

 

- 2 landing gear problems

 

- 1 smoke in cockpit

 

- 1 rudder cable broken

 

- 1 rudder damage (at VNE)

 

- 1 propeller blade detached

 

- 1 engine failed- no explanation

 

- 1 engine abnormal ops ( coolant hose- precautionary landing c/o)

 

I don't know enough about the reporting to say for sure, but I would reasonably assume (perhaps wrongly), that for a fatality to be reported because of "operational" issues, that it would suggest that it was likely that there wasn't obvious evidence of a mechanical or structural problem, or even that there was evidence that the way it was operated is what caused them to crash and subsequently die.

 

I have attached the ATSB file, but I imagine most are already aware of it.

 

- ultralight accidents 00-102.xlsx

 

- ultralight accidents 00-102.xlsx

 

- ultralight accidents 00-102.xlsx

[aro]

If you have anything more than a baseless opinion on which to condemn RAAus pilots, I reiterate that you should communicate it to the ops manager and CASA immediately; and also present the justification for your apparently baseless insistence that RAAus experimental homebuilts - and aircraft built to partial design standards, such as the Skyfox - are just as safe as Cessna 172s maintained in accordance with GA practise.

My ANALYSIS of the DATA is now fully on the table. Where is yours?

I'm not condemning RAAus pilots at all. My only point is that I believe your logic is damaged by too many assumptions. Your conclusions might be 100% correct, but you don't show the link.

 

You calculated the result if RAA had the same human factors accident rate as GA, and got an implausible result.

 

You calculated the result if RAA had the same hardware failure rate as GA, and got an implausible result.

 

From this, you concluded that the result was somewhere in the middle, and picked the number 3. Why not 2? 4? 1.5 (i.e. 50% higher)?

 

Isn't it possible that the data you are comparing are not directly comparable i.e. you are comparing apples with oranges? In fact, if you are comparing all GA I am fairly sure they are not directly comparable.

 

Your conclusion flows directly from the number you assumed, but you show no evidence to support this number, other than it is likely to be somewhere between 1 an 14.

 

I even have doubts about 14 being implausible. If the majority of GA hours are commercial, would it be that unlikely that GA charter has a 95% lower rate of pilot error accidents than RAAus? More information about the original data is required.

 

I don't have an analysis of the data. My official position is "I don't know". But you don't need to be a photographer to know if a photo is out of focus. Likewise, being out of focus doesn't mean it's not a photo of what the photographer says it is. But if you can't tell, the photo is no use.

 

 

[Bob Llewellyn]

I'm not condemning RAAus pilots at all. My only point is that I believe your logic is damaged by too many assumptions. Your conclusions might be 100% correct, but you don't show the link.You calculated the result if RAA had the same human factors accident rate as GA, and got an implausible result.

 

You calculated the result if RAA had the same hardware failure rate as GA, and got an implausible result.

 

From this, you concluded that the result was somewhere in the middle, and picked the number 3. Why not 2? 4? 1.5 (i.e. 50% higher)?

 

Isn't it possible that the data you are comparing are not directly comparable i.e. you are comparing apples with oranges? In fact, if you are comparing all GA I am fairly sure they are not directly comparable.

 

Your conclusion flows directly from the number you assumed, but you show no evidence to support this number, other than it is likely to be somewhere between 1 an 14.

 

I even have doubts about 14 being implausible. If the majority of GA hours are commercial, would it be that unlikely that GA charter has a 95% lower rate of pilot error accidents than RAAus? More information about the original data is required.

 

I don't have an analysis of the data. My official position is "I don't know". But you don't need to be a photographer to know if a photo is out of focus. Likewise, being out of focus doesn't mean it's not a photo of what the photographer says it is. But if you can't tell, the photo is no use.

We progress! Re-read my post, I explain why I think that it's reasonable to assume that RAAus pilots are not more than 3 times worse than GA at making decisions.

I would note that the majority of commercial operations outside of transport category, are lumped into the GA accident rate, and provided CAR 35s with quite a bit of work. On what basis do you think someone with a commercial rating is superhuman?

 

As to the comparability of the data: I firstly remind you that I'm comparing similar stages of development of regulation in the interests of safety, which requires comparing US GA in the '50s through early '70s to Australian ultralight data. Now, in the US in this period, single engined instrument trainers were few and far between; the majority of pilots operating GA were flying single engined aircraft in VFR, with a bit of night VMC thrown in. By the '70s radio aids were being found in single cockpits. Over the same period, the majority of single lighties changed from the "grasshoppers" and their civil derivatives, to the inimitable Cessna 172 / Beech Debonair / Piper Cherokee; the average stall speed crept up from 40~45kts to 52~57kts (at the top end of town), and constant speed propellors began making a showing.

 

Furthermore, most of the pilots playing GA over this period were not military trained, but the people too young for WW2.

 

So, it's pretty much apples with apples. Note too that the GA accident rate towards the end of the period was ~ twice the current US GA accident rate; the systemic immaturity had an effect.

 

With regard to your struggle with vastly varying HF accident rates, there is no point in further discussion until you can find your way around a normal distribution - I'm not trying to be offensive, but until you can appreciate the difference between 2 sigma and three sigma, you won't see the absurdity of considering 6 sigma as a realistic probability.

 

We know the photo is out of focus. We know that RAAus has, since HORSCOTS, focussed on pilot training. We know that the gross fatality RAAus accident rate has bettered the GA accident rate at a similar stage of development. We know that the RAAus accident rate has been getting worse for four years. We know that the comparable GA accident rate did not get worse over a similar period, or at any time subsequently (to a comparable degree). We know that GA had similar pilot training, indeed less developed. We know that GA had a strong maintenance system. We know that RAAus has nothing comparable as a maintenance system.

 

The differences here? RAAus accident rate getting worse; RAAus no comparable maintenance system. It behoves us all to either demonstrate, with a high level of confidence, that the RAAus accident rate increase has nothing to do with the ongoing lack of a maintenance system; or to improve the maintenance system. Just sitting there saying "I don't know but I disagree", or even "I don't know so I'll do nothing", are delinquent responses.

 

 

[Guest Nobody]

This is all great discussion and a no doubt maintinance is an issue but how will requiring endorsements for each aircraft as proposed help that?

 

 

[Bob Llewellyn]

This is all great discussion and a no doubt maintinance is an issue but how will requiring endorsements for each aircraft as proposed help that?

Not at all. I suggest that the statistics we do have, indicate that the only change that might benefit pilot training, is to return to the original AUF deal of spin training in gliders. But maintenance is, I believe, becoming the bleeding sore of Recreational Aviation safety at this point in time.

 

 

[aro]

We progress! Re-read my post, I explain why I think that it's reasonable to assume that RAAus pilots are not more than 3 times worse than GA at making decisions.

I would note that the majority of commercial operations outside of transport category, are lumped into the GA accident rate, and provided CAR 35s with quite a bit of work. On what basis do you think someone with a commercial rating is superhuman?

 

As to the comparability of the data: I firstly remind you that I'm comparing similar stages of development of regulation in the interests of safety, which requires comparing US GA in the '50s through early '70s to Australian ultralight data. Now, in the US in this period, single engined instrument trainers were few and far between; the majority of pilots operating GA were flying single engined aircraft in VFR, with a bit of night VMC thrown in. By the '70s radio aids were being found in single cockpits. Over the same period, the majority of single lighties changed from the "grasshoppers" and their civil derivatives, to the inimitable Cessna 172 / Beech Debonair / Piper Cherokee; the average stall speed crept up from 40~45kts to 52~57kts (at the top end of town), and constant speed propellors began making a showing.

 

Furthermore, most of the pilots playing GA over this period were not military trained, but the people too young for WW2.

 

So, it's pretty much apples with apples. Note too that the GA accident rate towards the end of the period was ~ twice the current US GA accident rate; the systemic immaturity had an effect.

 

With regard to your struggle with vastly varying HF accident rates, there is no point in further discussion until you can find your way around a normal distribution - I'm not trying to be offensive, but until you can appreciate the difference between 2 sigma and three sigma, you won't see the absurdity of considering 6 sigma as a realistic probability.

 

We know the photo is out of focus. We know that RAAus has, since HORSCOTS, focussed on pilot training. We know that the gross fatality RAAus accident rate has bettered the GA accident rate at a similar stage of development. We know that the RAAus accident rate has been getting worse for four years. We know that the comparable GA accident rate did not get worse over a similar period, or at any time subsequently (to a comparable degree). We know that GA had similar pilot training, indeed less developed. We know that GA had a strong maintenance system. We know that RAAus has nothing comparable as a maintenance system.

 

The differences here? RAAus accident rate getting worse; RAAus no comparable maintenance system. It behoves us all to either demonstrate, with a high level of confidence, that the RAAus accident rate increase has nothing to do with the ongoing lack of a maintenance system; or to improve the maintenance system. Just sitting there saying "I don't know but I disagree", or even "I don't know so I'll do nothing", are delinquent responses.

I don't think people with commercial ratings are superhuman, but we know that commercial operators are supposed to have standard operating procedures etc. as part of an AOC, and I would be astonished if they didn't have an accident rate much better than 3x less than RAAus.

 

You are comparing comparing US GA in the '50s through early '70s to Australian ultralight data. You say they are directly comparable, but what evidence do you have? You should be able to produce some statistics to demonstrate they are more comparable than RAAus vs. GA today, for example. I personally don't buy it.

 

I am not unfamiliar with a normal distribution. I would make a few couple of points:

 

- It only make sense to fit values into a distribution if they are part of the same population. There is evidence in this case they are not, e.g. if you assume the HF accident rate, or the maintenance problem rate are the same, the results are implausible. This to me is evidence the populations are not the same.

 

- 2 sigma, 3 sigma, 6 sigma - being outside these bounds is not impossible, it just means that a value is not likely to be part of the originally measured population i.e. there is something different about it.

 

- I think you are confusing the standard deviation and the mean. I have not referred to the standard deviation, and I think this is the first time you have mentioned it. A value 6x the mean is not remarkable in itself.

 

Asking for evidence is not a delinquent response. It is part of ensuring that we don't waste our time and energy in one area when the problem lies in another.

 

 

[dodo]

This thread was about a requirement for pilot training for each type.

 

Maybe another thread could be used to discuss RA vs GA maintenance standards?

 

dodo

 

 

[M61A1]

This thread was about a requirement for pilot training for each type.Maybe another thread could be used to discuss RA vs GA maintenance standards?

 

dodo

If what I am told is correct, RAA maintenance is another issue that will see some change, so it may be a little more relevant than you thought.

 

 

[dodo]

It may be relevant to RA, and it may be fascinating, but it isn't relevant to the thread.

 

dodo

 

 

[M61A1]

It may be relevant to RA, and it may be fascinating, but it isn't relevant to the thread.

dodo

Not for type training, but for Ops manual changes it may be.

 

 

[Bob Llewellyn]

I don't think people with commercial ratings are superhuman, but we know that commercial operators are supposed to have standard operating procedures etc. as part of an AOC, and I would be astonished if they didn't have an accident rate much better than 3x less than RAAus.You are comparing comparing US GA in the '50s through early '70s to Australian ultralight data. You say they are directly comparable, but what evidence do you have? You should be able to produce some statistics to demonstrate they are more comparable than RAAus vs. GA today, for example. I personally don't buy it.

 

I am not unfamiliar with a normal distribution. I would make a few couple of points:

 

- It only make sense to fit values into a distribution if they are part of the same population. There is evidence in this case they are not, e.g. if you assume the HF accident rate, or the maintenance problem rate are the same, the results are implausible. This to me is evidence the populations are not the same.

 

- 2 sigma, 3 sigma, 6 sigma - being outside these bounds is not impossible, it just means that a value is not likely to be part of the originally measured population i.e. there is something different about it.

 

- I think you are confusing the standard deviation and the mean. I have not referred to the standard deviation, and I think this is the first time you have mentioned it. A value 6x the mean is not remarkable in itself.

 

Asking for evidence is not a delinquent response. It is part of ensuring that we don't waste our time and energy in one area when the problem lies in another.

Ok, so you have nothing tangible on commercial operator's accident rates vs RAAus, you just think my arbitrary "no more than 3 times GA average" should be "no more than much more than 3 times charter average". So what?

Ok, you personally don't buy Airworthiness Regs or Pilot Training as relevant to accident rates. You're entitled to that opinion.

 

No, you are arguing that the two populations have a vastly different skewness or Weibullness, which there is no evidence whatsoever to support.

 

It is not the first mention I have made of the standard deviation; and as I understand maths, the occurence of an event with probability of >1 / 99.9999 is much more remarkable than that of an event with probability 1 / 99.7.

 

You have so far rejected any relationship between Airworthiness Regs or Pilot Training and accident rate; you have rejected any similarity between US GA of the 1950-1970s and RAAus operations; you have rejected any similarity between Piper Colts and Cubs, Cessna 140s and 170/172s, etc, and RAAus aeroplanes; you have rejected any similarity between US air and Australian air; AND YOU HAVE GIVEN NO BASIS FOR THESE REJECTIONS.

 

There are none so blind as those that will not see.

 

 

[fly_tornado]

Interesting, probably makes sense for those that have really only ever flown one type of aircraft. Someone that has only ever flown Tecnam's are going to be a bit interesting in a Jabiru and so I'd encourage this. As long as instructors treat it right and don't rip people off by it.

It sounds like a good money earner and not much else. Where is the evidence that pilots are killing themselves in the first 5 hours of flying a new type of aircraft though? They might be crashing because the certificate is letting pilots fly with inadequate training.

 

 

[Dafydd Llewellyn]

When you go to hire an aircraft, no operator will simply hand you the keys unless he's satisfied you are competent to fly it. So if you go the hire a Jabiru, with a background of only Tecnam experience, or vice-versa, you are going to get put through the hoops anyway. Formalising this by way of an endorsement that has to be processed by RAA is simply useless make-work. Sure, you can go buy one, but the dealer will normally provide some conversion experience; he has an insurance premium to consider, too.

 

It's a bloody fool idea.

 

 

[motzartmerv]

I really wonder what the thinking is behind this rule change. Whats CASA's line in the RPL? Just spitballing, but could ICAO have something on this that CASA (and the RAA) are trying to adopt? (re adopt).

 

I know the HP, LP thing is gone from the new ops manual. Perhaps this is an attempt to take account for pilots transfering from say a Jab to a drifter or thruster. At the moment they need a LP endo. Maybe this is the RAA's way of dealing with the obvious problems this could cause?.

 

I dunno. i just think there has to be some reason behind the change. Surely!!!

 

 

[Oscar]

Seems to me there is a very considerable risk of a major unintended consequence of this move, being one of potential liability on the part of the person who provides the type endorsement. As things stand, (if I am one the right track here), Instructors have some level of protection in the form of the Training Syllabus, which provides a standard against which the Instructor' training can be measured. That in turn provides at least the basis of a defence in law that the training provided was in accordance with a requirement, so the Instructor's liability isn't open-ended that they have failed to provide required training in aspects of flying competency that may have 'caused' an accident. (Of course, that's not to say that there aren't lawyers who will try it on, but it does at least give Instructors some sort of base-line for a defence.)

 

Surely, short of having an 'approved' standard for type endorsement for each aircraft, the level of exposure of risk for Instructors who provide type endorsements could become very considerably higher. I am pretty confident that any good Instructor will in fact provide an endorsee with all the information and testing they believe necessary for such an endorsement - but that would be a matter of their personal judgement rather than 'ticking off a list' that is approved. Once personal judgement is involved, it opens up a pretty vast avenue of attack for the legal fraternity.

 

Imagine the chaos of this type of requirement for motor vehicles. You have learned to drive in a vehicle with ABS and ESC systems, let's say a modern Falcon or Holden. You want to drive, let's say, an early classic Porsche 911. In theory at least, an Instructor would be considered to have failed to give adequate training for an endorsement without going through skid and spin recovery and braking techniques. Ok, any Instructor worth her or his salt would know that at least covering the basics of that is pretty much survival-level stuff for an early 911 - but unless there is a 'syllabus', there is very little cover for the Instructor to say 'I gave the training that was necessary'.

 

Now, manufacturers/agents for 24-reg aircraft might well be able to develop such a 'syllabus' for their machines and they would be responsible for the accuracy and completeness of that syllabus. Where does that leave 19-reg machines?

 

It seems to me that there is an element in a 'check flight' of mutual acceptance of risk, i.e. the check-flight person needs to be convinced that the checkee is adequately competent in handling the thing to be able to fly it safely and the checkee accepts that the information/experience is sufficient introduction to the aircraft in question. I think there's a fairly large jump from that to the apparent liability an Instructor may have if he/she signs a specific 'type endorsement'.