Instrument Rating Test in Highly Automated Aircraft

[Tee Emm]

The British CAA have recently published a paper called CAA Paper 2011/03 - CAA`Significant Seven`Task Force Reports.

Link: http://www.caa.co.uk/docs/33/2011_03.pdf

It deals with the seven most common causes of aircraft accidents and recommendations to reduce them.

Report No:1, deals with Loss of Control which is the current leading cause of accidents in jet transports. Automation comes into the picture here.

One paragraph is interesting in the Australian context of command instrument rating tests on jet transports. It states:


Currently, when testing and checking pilots’ competence, the principal focus is on their handling skills rather than their monitoring skills.

In Australia, there is anecdotal evidence that most instrument approach procedures during airline instrument rating tests are with full use of automatics. This includes such things as engine failure immediately after take off where manufacturers advise the engaging of the automatic pilot as soon as practicable - within operating limitations. During the conduct of an instrument rating test in (say) a B777, A320 or B737, the principle focus is on the early engagement of and monitoring of the automatic pilot and its associated modes. In fact, the actual hand flying non-automatics (raw data) part of the IR test would normally cover less than ten percent of the test.
It seems strange therefore the CAA report would claim that their research indicates the principle focus of a test (in highly automated aircraft) is on pilot handling skills and not autopilot monitoring skills. The opposite seems to be true in Australia.

[greenslopes]

Quit true, LOCI is the largest statistical cause of not only hull losses but also onboard and on ground fatalities.
Unfortunately the reasons for LOCI are many and varied and embrace most core elements of the human factors spectrum.
Boeing and Airbus have each completed studies looking for primary factors and have identified over-reliance on automatics combined with a reduction in basic flying skills(due lack of use).
Where to from here for the regulators?

[Feather #3]

My major beef with the current emphasis in Australia [allegedly following manufacturer's procedures and World's Best Practice!] is that nobody is training for what happens when the autopilot fails!!:uhoh:

And they do!

G'day ;)

[waren9]

Or looking at it from the other direction, how many loss of control accidents have Australian operators had? How many prangs due to A/P failure? How many jet transports in Australia only have 1 A/P?

Use of the automatics is a whole skillset itself.

Put another way, how many prangs have inappropriate use of the automatics as a significant contributing factor? I think many.

Just have to spend 5 minutes in the Teck Log forum see how how widely misunderstood some systems are. Especially Airbus automatics.

[A37575]

Boeing and Airbus have each completed studies looking for primary factors and have identified over-reliance on automatics combined with a reduction in basic flying skills(due lack of use).


That has been known for years so the study is nothing new. The obvious fix is to ensure basic flying skills are practiced in the simulator and until that happens the problem will not go away. 15 minutes of hand flying using a flight director in a two hour simulator period, is worth SFA in terms of improving basic flying skills.

[The Green Goblin]

Perhaps to play devils advocate, 2 tests need to be passed. An automatic test, and a manual flying test for each instrument renewal.

GG

[Capn Bloggs]

To quote from the CAA's paper:

Recommendation 5 states that: “The CAA, or better still EASA, should establish a working group that draws on the training expertise within the UK airline industry and other organisations, to identify best practices in human flight deck monitoring and to propose training methods for the same.

A possible output from such a group could be a DVD training aid to highlight awareness and suggest training strategies.” EASA, even if it were to accept the recommendation, would be unlikely to act upon it in the short term. Therefore, it is proposed that the task of setting up a working group from the UK aviation industry (Phase 1 of this proposal) be allocated to the CAA Human Factors Programme Manager (who, incidentally, was a member of the LoCTF).

The output from this working group would lead to Phase 2:
“Operators should adopt strategies for training effective monitoring skills. Pilots must develop techniques for monitoring the performance and behaviour of the other pilot, the operation of the aircraft (e.g. appropriate mode selections), the control of the flight path and aircraft energy, communications, task management, and situational awareness. Particular emphasis should be placed on the pilots checking that the autoflight systems are controlling the aircraft correctly. They must develop effective intervention strategies (when and how to intervene) and the concept of the three levels at which errors can be trapped: (1) avoidance by anticipation, (2) recognition before occurrence, and (3) early and effective mitigation.”

Q: The $64k is how on earth do you train pilots to monitor properly?
A: Make them FLY the aeroplane regularly so their eyes are trained and practised to know where to look. The same cross-reference technique is used whether flying or monitoring. Make them do a manual descent occasionally so they are very familiar with where the aeroplane should be during descent.

[Nulli Secundus]

I remember a checkie who always encouraged us to hand fly as much as possible. He also recommended disconnecting the A/P as early as possible during the approach, IMC or VMC, so that a) you had a better feel of the aircraft & conditions prior to minima & b) to help keep skill levels up. Regardless of schedule, his advice was to fly a practice approach at least once a week.

[A37575]

I remember a checkie who always encouraged us to hand fly as much as possible

And the case of the Virgin Blue F/O who asked the captain if it was OK to manually fly the 737 from 15,000 ft down. Weather CAVOK which is why he asked the captain. The captain gave the OK but warned the F/O he must keep the flight director on. CAVOK - and still the experienced captain wanted the crutch of the FD. Is it any wonder that F/O's give up trying to maintain their flying skills?

[compressor stall]

My major beef with the current emphasis in Australia [allegedly following manufacturer's procedures and World's Best Practice!] is that nobody is training for what happens when the autopilot fails!!

It's a balance.

I was told of an engine failure at an overseas airline where the pilots disconnected the autopilot and flew it manually as that was how every sim session had been done - engine failure and subsequent simultaneous autopilot failure. The crew were more comfortable flying manually.

Negative training...

[lederhosen]

On a recent check I spent about 15 minutes in one scenario manually flying including a trim runaway, followed by an engine overheat / shutdown while the copilot worked his way through the qrh. I was a bit sceptical given the Boeing single failure guidelines. But I must say that it was good practice, with the aircraft pretty much at its performance limit. It seemed to make the set piece V1 cut and single engine raw data easier too.

[Bankstown Boy]

In someways this is outside my area of expertise but in someways not. After all we are all doing the same thing here. My daily ride is a piston twin and what passed for automatics from those days.

However, I believe that the same points are valid here. When I fly, in the real world, the autopilot flies the plane and I manage the process and procedures. During annual cmd renewals, I am allowed one of the three approaches on autopilot but the other two must be manually flown.

This is not machoism, it is a practical renewal of skills. Whilst I will readily admit that my technology is not as sophisticated or as dependable as that available to even the average airline pilot, the point is that all systems can and do fail, generally at the most inappropriate time.

It is our job, at that point, to actually be able to pilot the aircraft. On this point I totally agree with the GG

[psycho joe]

Bankstown Boy

That's all very well, but the aircraft that I fly has so many levels of redundancy, that the amount of failures necessary for me to resort to raw data manual flying is so statistically improbable that you'd sooner win lotto and simultaneously be struck by lightning whilst riding down the main street buck naked on a pantomime horse.

...Yet when you get into a simulator :rolleyes: