The lure of automation
Moderator
Thread Starter
The lure of automation
I am a VFR pilot, with lots of long expired IFR flying skills - it's been about 30 years since I flew hard IFR, but back then I did quite a bit, and all by hand. In those days, it was ADF, and ILS. DME was a luxury, and LORAN and GPS had yet to come. None of the planes I flew had RNav, so I never got to know that. I'm happy being a good, current VFR pilot, with a resource of skills for poor conditions, but the desire to avoid them.
Automation, even in GA aircraft seems like a nice warm blanket of security of capability and situational awareness - and in some cases a trap! The magenta line is not always the best route!
Lat month, I was flying the ultimately well equipped C 182, with full Garmin glass cockpit. The owner and I were flying south through Finland, in VFR conditions which were not great. I had elected left seat for that flight, knowing that in more poor conditions, I wanted everything going for me. We knew we would encounter low weather, but it was uncertain if it would be before our intended destination, so we set out, with an alternate along the way, just in case.
The weather went down to the low limits of VFR just short of our destination - Olou. No problem, I would back track to Kemi, my alternate. But Kemi just closed. Olou accepted my request for special VFR, so I continued on to Olou. With all this glass cockpit at my disposal, how could it not go well?
Most of the Garmin was still set to the factory settings, as we're still "getting used to it", and customizing to our needs. Placing a magenta line from my present position to Olou was not a problem. Flying that line would not be a problem. But flying it at 700 feet made me nervous. I could climb, but that would violate the spirit of special VFR, and really amount to continued VFR into IFR conditions. I'm too old for that silliness...
What I could not figure our how to do was to reset the Garmin to have the magenta line extend out the runway centerline so I could intercept it over the Baltic Sea, where I knew I was safe at 700 feet. So, I reverted to what I knew from instinct, the localizer on NavCom 2. I was much more certain that there would not be windmills on the extended centerline of the runway. The Magenta line became not much more that a distraction, as I established on the localizer, when waited to intercept the glideslope from below. Once on the localizer, I did reset the GPS, but by then, it was nothing more than situational awareness.
So far, all that glass cockpit, situational awareness had not provided meaningful obstacle awareness, and I could not understand why. Once I was within five miles of the airport, and I zoomed in, it did - windmills on both sides of me! I happily saw the runway lights two miles back, at about 500 feet, so no problem, and could have easily flown a lower approach than that if needed. The next morning, I took off the reciprocal, and was quite aware of many windmills on either side of the approach path. They did not reach to my 700 feet of the previous evening, but I sure was happy I had not ducked down to see where I was going!
After some menu searching while right seat, I found that the factory setting to of the GPS did not provide obstacle display if the zoom was more than 5 miles. I reset that to 50 miles, and a whole vista of hazards presented itself!
I learned to ignore the automation when things get tense, revert to the simple things which you know. It's a sad commentary that the aircraft possessed immense capability, and I could not use it, but trying to learn it in the difficult conditions would be even more dangerous!
This aircraft, though highly automated, is probably still a "basic" plane, compared to the automation found in other aircraft. But the theme is the same, there comes a point where all that automation can change from being an asset to a lure and hazard - you must maintain the basic skills to safely fly without it, and then the additional skill to prevent that automation luring you farther than your basic skills can handle.
I learned needle, ball and airspeed, with a map, and looking at the ground. Everything since then has been new learning for me, but the old skills reside in my instinct. I worry for newer pilots, learning in an environment, where there is so much more technology, but seemingly little more training time allocated to training that on top of the basic skills. Are both sets of skills being mastered during training? 'Cause they're not the same skills!
Automation, even in GA aircraft seems like a nice warm blanket of security of capability and situational awareness - and in some cases a trap! The magenta line is not always the best route!
Lat month, I was flying the ultimately well equipped C 182, with full Garmin glass cockpit. The owner and I were flying south through Finland, in VFR conditions which were not great. I had elected left seat for that flight, knowing that in more poor conditions, I wanted everything going for me. We knew we would encounter low weather, but it was uncertain if it would be before our intended destination, so we set out, with an alternate along the way, just in case.
The weather went down to the low limits of VFR just short of our destination - Olou. No problem, I would back track to Kemi, my alternate. But Kemi just closed. Olou accepted my request for special VFR, so I continued on to Olou. With all this glass cockpit at my disposal, how could it not go well?
Most of the Garmin was still set to the factory settings, as we're still "getting used to it", and customizing to our needs. Placing a magenta line from my present position to Olou was not a problem. Flying that line would not be a problem. But flying it at 700 feet made me nervous. I could climb, but that would violate the spirit of special VFR, and really amount to continued VFR into IFR conditions. I'm too old for that silliness...
What I could not figure our how to do was to reset the Garmin to have the magenta line extend out the runway centerline so I could intercept it over the Baltic Sea, where I knew I was safe at 700 feet. So, I reverted to what I knew from instinct, the localizer on NavCom 2. I was much more certain that there would not be windmills on the extended centerline of the runway. The Magenta line became not much more that a distraction, as I established on the localizer, when waited to intercept the glideslope from below. Once on the localizer, I did reset the GPS, but by then, it was nothing more than situational awareness.
So far, all that glass cockpit, situational awareness had not provided meaningful obstacle awareness, and I could not understand why. Once I was within five miles of the airport, and I zoomed in, it did - windmills on both sides of me! I happily saw the runway lights two miles back, at about 500 feet, so no problem, and could have easily flown a lower approach than that if needed. The next morning, I took off the reciprocal, and was quite aware of many windmills on either side of the approach path. They did not reach to my 700 feet of the previous evening, but I sure was happy I had not ducked down to see where I was going!
After some menu searching while right seat, I found that the factory setting to of the GPS did not provide obstacle display if the zoom was more than 5 miles. I reset that to 50 miles, and a whole vista of hazards presented itself!
I learned to ignore the automation when things get tense, revert to the simple things which you know. It's a sad commentary that the aircraft possessed immense capability, and I could not use it, but trying to learn it in the difficult conditions would be even more dangerous!
This aircraft, though highly automated, is probably still a "basic" plane, compared to the automation found in other aircraft. But the theme is the same, there comes a point where all that automation can change from being an asset to a lure and hazard - you must maintain the basic skills to safely fly without it, and then the additional skill to prevent that automation luring you farther than your basic skills can handle.
I learned needle, ball and airspeed, with a map, and looking at the ground. Everything since then has been new learning for me, but the old skills reside in my instinct. I worry for newer pilots, learning in an environment, where there is so much more technology, but seemingly little more training time allocated to training that on top of the basic skills. Are both sets of skills being mastered during training? 'Cause they're not the same skills!
Are both sets of skills being mastered during training? 'Cause they're not the same skills!
A typical airline type instrument rating in Australia is 90% automatics although the rules say "at least one instrument approach must be flown without the autopilot or flight director." If regulators were serious about the potential dangers of automation addiction instead of simply paying lip service, they would require that 50% of the test should be manually flown. Unfortunately that would probably result in a high rate of failures and that costs money.
Join Date: Jun 2015
Location: Soon to be out of the EU.
Posts: 0
Likes: 0
Received 0 Likes
on
0 Posts
One of the Airbus golden rules is to us an 'appropriate' level of automation for the task in hand. When things don't do as planned, intervene and don't just sit for automatics to sort themselves out.
The children of the magenta line video is outdated. Things (in most parts) have moved on and lessons have been learned. However on watching it please don't think that we airline pilots sick there playing angry birds letting the computer work it's magic which some watchers of this video often proclaim online without fully knowing the facts.
It's fatigue and landing long for landing from an unstable approach that people should be more worried about these days but automation dependency will likely cause more issues in problem.
The children of the magenta line video is outdated. Things (in most parts) have moved on and lessons have been learned. However on watching it please don't think that we airline pilots sick there playing angry birds letting the computer work it's magic which some watchers of this video often proclaim online without fully knowing the facts.
It's fatigue and landing long for landing from an unstable approach that people should be more worried about these days but automation dependency will likely cause more issues in problem.
Last edited by HeartyMeatballs; 20th Jul 2015 at 16:58.
The children of the magenta line video is outdated
Instead of immediately turning to track in the required direction to that aid, in most cases the pilot's immediate action is to continue on the original track before the ATC request, and go heads down into the CDU and type in the navaid so that L NAV can be used to fly the 25 miles. Meanwhile the aircraft flies another seven miles away from the aid before the final button pressing is completed and only then does the aircraft begin to turn.
I recall reading a 737 accident report from the Middle East where after take off and during the initial climb at night, the captain had difficulty engaging the autopilot. He thought it was engaged but it wasn't. Subsequently the aircraft gradually went into a spiral dive and never recovered. The last words recorded on the CVR was the captain screaming to his first officer to engage the autopilot despite the 737 being now in a fatal unusual attitude. That is automation dependency to the bitter end
Join Date: Jan 2008
Location: Canada
Posts: 631
Likes: 0
Received 0 Likes
on
0 Posts
I just rewatched "Children of Magenta". I found it to be very worthwhile. Though nearly none of my flying is highly automated, it is reassuring to be reminded that we must fly the plane first. Automation is less the solution to an unusual event, it is more an assistant to normal operations.
I recommend the video...
I recommend the video...
Join Date: Nov 2000
Location: Cambridge, England, EU
Posts: 3,443
Likes: 0
Received 1 Like
on
1 Post
I just rewatched "Children of Magenta".
The problem may have started long before take off:
Most of the Garmin was still set to the factory settings, as we're still "getting used to it", and customizing to our needs.............What I could not figure out how to do was to reset the Garmin to....... I could not understand why.............After some menu searching while right seat, I found that the factory setting to of the GPS did not provide obstacle display....
Moderator
Thread Starter
Yes, indeed... I failed to familiarize myself with software architecture with appeared to be at least very similar to the Garmin Aera 500 I have been happily using for years, and the G 1000 I have flown many times. As I fly only VFR, and usually in a locale I know well, I have never needed to use these very advanced systems for any purpose other than back up situational awareness.
When I compare the terrain and obstacle presentation of the GTN 750 and the Aera 500, they appear to be the same, but unless you change default settings in the 750, it does not function as the 500 does. I am far from a software wiz, so I spent a lot of time reading the manual, both before the flying, and during, while I was right seat. However, in hind sight some of the differences which I later found to be rather important to me, did not leap off the page at me. I did receive two hour hour training sessions of the Garmin systems in this aircraft, one with the installed, and a second with a visiting Garmin rep. That left me feeling fairly confident about my ability to use the system in a VFR environment. But, although I have been evaluated and found competent to fly very complex GA aircraft, there seems no mechanism to assure my competence with an even more complex optional system in these aircraft.
This is where the lure of automation begins (for me anyway). I am presented with what appears to be an incredibly capable system, without the caution that to receive the full benefit of that capability, I'd have to reset a number of default settings. There does not seem to be a "checklist" to be fulfilled to assure that the important elements of the system are understood by the pilot, for the environment in which they will fly. This is a miss....
The operator's manual for the GTN 750 has more content than several Cessna flight manuals combined! We have pilots who fly planes without reading and understanding the whole flight manual, so understanding and being able to recall the contents of a many hundred page system manual is not assured.
For me, two misunderstandings took place, which were relevant to my failing: Unlike the Aera 500, which displays obstructions out to a 50 mile page zoom, the GTN 750 was preset to stop displaying them beyond a 5 mile page zoom. So I'm trying to optimize my situational awareness in a 20 mile approach area, and for lack of understanding this function well, I cannot see the airport and obstructions at the same time. But nothing prompted me to know this, other than my realizing "some wasn't right". Approach in poor conditions was not the time to figure out why, so I stopped placing any dependance upon it.
Similarly, the Aera 500 will allow me to "draw" a waypoint between my present position, and same destination. The GTN 750 will not. It has an "Insert Waypoint" function, but after trying and failing, and later reading the book for that, it turns out that the functionality is not the same as the Aera 500, so I was doomed to fail at what I was attempting.
So, as my primary role was mentoring my friend in his plane, I spent literally hours optimizing the settings of the units to his needs, and briefing him in great detail as to what I had set, and why, and how he should use it. I explained what not to attempt (appropriate to his experience level), and that he is to ignore the system if he becomes task saturated. He is not a good English reader, so I could not depend that he would read and understand the operator's manual for these systems.
In the final analysis, when I could most have benefited from the additional capability of the automation in this aircraft, I relegated it to low level advisory, while I returned to much simpler systems, I knew very well, and it worked.
There will be a future shift, where new pilots will devote the immense effort required to be conversant with these automated systems as a part of their basic pilot training. During this process, will they become conversant with the differing software architectures of different manufacturers? Will they retain the much more basic skills to continue an adequately safe flight using only the most basic backup systems (needle ball and airspeed, VOR, ADF and localizer)?
When I compare the terrain and obstacle presentation of the GTN 750 and the Aera 500, they appear to be the same, but unless you change default settings in the 750, it does not function as the 500 does. I am far from a software wiz, so I spent a lot of time reading the manual, both before the flying, and during, while I was right seat. However, in hind sight some of the differences which I later found to be rather important to me, did not leap off the page at me. I did receive two hour hour training sessions of the Garmin systems in this aircraft, one with the installed, and a second with a visiting Garmin rep. That left me feeling fairly confident about my ability to use the system in a VFR environment. But, although I have been evaluated and found competent to fly very complex GA aircraft, there seems no mechanism to assure my competence with an even more complex optional system in these aircraft.
This is where the lure of automation begins (for me anyway). I am presented with what appears to be an incredibly capable system, without the caution that to receive the full benefit of that capability, I'd have to reset a number of default settings. There does not seem to be a "checklist" to be fulfilled to assure that the important elements of the system are understood by the pilot, for the environment in which they will fly. This is a miss....
The operator's manual for the GTN 750 has more content than several Cessna flight manuals combined! We have pilots who fly planes without reading and understanding the whole flight manual, so understanding and being able to recall the contents of a many hundred page system manual is not assured.
For me, two misunderstandings took place, which were relevant to my failing: Unlike the Aera 500, which displays obstructions out to a 50 mile page zoom, the GTN 750 was preset to stop displaying them beyond a 5 mile page zoom. So I'm trying to optimize my situational awareness in a 20 mile approach area, and for lack of understanding this function well, I cannot see the airport and obstructions at the same time. But nothing prompted me to know this, other than my realizing "some wasn't right". Approach in poor conditions was not the time to figure out why, so I stopped placing any dependance upon it.
Similarly, the Aera 500 will allow me to "draw" a waypoint between my present position, and same destination. The GTN 750 will not. It has an "Insert Waypoint" function, but after trying and failing, and later reading the book for that, it turns out that the functionality is not the same as the Aera 500, so I was doomed to fail at what I was attempting.
So, as my primary role was mentoring my friend in his plane, I spent literally hours optimizing the settings of the units to his needs, and briefing him in great detail as to what I had set, and why, and how he should use it. I explained what not to attempt (appropriate to his experience level), and that he is to ignore the system if he becomes task saturated. He is not a good English reader, so I could not depend that he would read and understand the operator's manual for these systems.
In the final analysis, when I could most have benefited from the additional capability of the automation in this aircraft, I relegated it to low level advisory, while I returned to much simpler systems, I knew very well, and it worked.
There will be a future shift, where new pilots will devote the immense effort required to be conversant with these automated systems as a part of their basic pilot training. During this process, will they become conversant with the differing software architectures of different manufacturers? Will they retain the much more basic skills to continue an adequately safe flight using only the most basic backup systems (needle ball and airspeed, VOR, ADF and localizer)?
Flight Directors - A Fatal Attraction?
My first encounter with flight directors was in 1966 while undergoing conversion to the Avro 748. The RAAF had seen fit to send me to Woodford in Cheshire, all the way from Australia to ferry the second of several new 748’s for the RAAF VIP squadron at Canberra. The conversion was conducted on a battered 748 demonstrator G-ARAY known as Gary. The contract allowed four hours of dual for the captains and nothing for the co-pilots.
G-ARAY had the basic instrument flying panel of that era and no flight director. Our instructors at Avro’s were well known test pilots Bill Else, Tony Blackman and Eric Franklin. Jimmy Harrison was chief test pilot. Unlike the bog-standard civilian 748, the RAAF 748’s were to be equipped with a Collins FD 108 FD. So the situation existed that the RAAF 748’s had a British Smith’s autopilot system which was married (somewhat expensively and painfully) to the American Collins FD 108. For the life of me I could not see why a flight director was needed in the RAAF 748. After all, the approach speed was that of a DC3 – 80 knots and the aircraft a delight to handle compared with the venerable Dak.
In retrospect, I think the old Wing Commander Transport Ops at Department of Air, who was charged with the procurement of the 748 for RAAF service, and hadn’t flown for years, was perhaps conned by the Avro sales people in conjunction with Collins, into buying the Collins systems. Certainly in my view as the squadron QFI, flight directors were not operationally needed. In the event, the RAAF machines came with Collins FD 108 flight directors and as the contract specified each captain would be given only one hour of dual instruction once the 748 came out of the factory, we needed to learn how to operate the FD.
First, a course was arranged at the Collins establishment at Weybridge in Surrey. The two RAAF captains and their co-pilots attended and our two navigators and our instrument fitters also turned up to enjoy the Collins hospitality. We learned about 45 degree automatic intercepts of the VOR and ILS beams and other goodies including V-bar interpretation. We were showered with glossy brochures of the flight director by white dust-coated lecturers and shown a film.
By lunch time the presentation was complete and we were shouted to a slap up pub meal with lots of grog, all paid for by Collins. We asked what further lectures were to take place after lunch. We were told the course was over – it was just a morning’s job and we were free to leave unless we would like more drinks. Naturally it was churlish to refuse and hours later we staggered to the railway station (I think), smashed to the eye balls and having forgotten all about the marvels of 45 degree auto intercepts on the FD 108. I must say it was a bloody good three hour course what with the free grog and all that.
A few weeks later, I flew the second RAAF aircraft out of the factory, A10-596, under the watchful eye of Eric Franklin DFC and he demonstrated flight director stuff. For example to climb using the FD, you first put the aircraft into a normal climb and when settled you switched on the FD and carefully wound up the pitch knob so that the little aeroplane sat in the middle of the V-bars.
I quickly realised that you hand flew the basic artificial horizon to whatever attitude was appropriate for the manoeuvre then told the FD 108 V bars where you wanted them. The ILS intercept of 45 degrees was never used because radar vectors didn’t do such angles. I became more and more convinced the 748 didn’t need flight directors and that they were a load of bollocks in that type of low speed aircraft. We were told the USAF used the FD 108 in its F4 Phantoms and that Collins was anxious to makes sales in the UK market. The RAAF Wing Commander got sucked in by good sales talk and from then on all RAAF 748’s became so equipped.
I held personal doubts about the usefulness of flight directors in general as I could see even then, their extended use could lead to degradation of pure instrument flying skills. Today’s flight director systems are light years ahead in sophistication compared with the old Collins FD 105 and 108 series. But the problem with blind reliance on FD indications and thus steady degradation of manual instrument flying skills is as real now as it was back in 1966.
Now to the present day - although first some background history. First published in 1967, “Handling the Big Jets” written by the then British Air Registration Board’s chief test pilot David Davies, is still considered by some as the finest treatise still around on jet transport handling. Indeed, the book was described by IFALPA as `the best of its kind in the world`…written by a test pilot for airline pilots, the book is likely to become a standard text book…particularly recommended to all airline pilots who fly jets in the future…valuable to those pilots who are active in air safety work”.
All that was back in 1967 and little has changed since then - apart from an increasing propensity for crashes involving loss of control rather than simply running into hills. LOC instead of CFIT. Mostly these accidents were caused primarily by poor hand flying and instrument flying skills which certainly explains why aircraft manufacturers lead the push for more and more automatics. A colleague involved with Boeing 787 training was told by a test pilot on type, that the 787 design philosophy was based on the premise that incompetent crews would be flying the aircraft and that its sophisticated automatic protection systems were in place to defend against incompetent handling. Be it a tongue-in-cheek observation, it contains an element of truth.
With the plethora of inexperienced low hour cadet pilots going directly into the second in command seats in many airlines in Asia, the Middle East and Europe, these protection systems are important.
Towards the end of his book, David Davies discusses the limitations of the flight instruments in turbulence and in particular the generally small size of the active part of the basic attitude information or the ‘little aeroplane” as many older pilots will remember it. He continues: “the preponderance of flight director and other information suppresses the attitude information and makes it difficult to get at” and “the inability, where pitch and roll information is split, to convey true attitude information at large pitch and roll angles in combination” Finally Davies exhorts airline pilots “not to become lazy in your professional lives…the autopilot is a great comfort, so is the flight director and approach coupler…but do not get into the position where you need these devices to complete a flight”. There is more but go and read the book.
Having done the unforgiveable and quoted freely from an eminent authority, it is time for Centaurus to say something original and accept the no doubt critical comment that is freely available on Pprune. Flight Directors can be a fatal attraction to those pilots who have been brain-washed by their training system to rely on them at all times. While Boeing in their FCTM advise pilots to ensure flight director modes are selected for the desired manoeuvre, it also makes the point that the FD should be turned off, if commands are not to be followed.
Recently a new pilot to the Boeing 737 asked his line training captain if he could turn off the FD during a visual climb so he could better “see” the climb attitude. His request was refused as being “unsafe” and instead he was told to “look through” the FD. I don’t know about you, but I find it impossible to “see” the little aeroplane when it is obscured by twin needles or V-bars. In fact it takes a fair amount of imagination and concentration to do so. Which may be why Boeing recommends pilots to switch off the FD if commands are not to be followed.
I well recall my first simulator experience in the 737 of an engine failure at V2 where I was having a devil of a time trying to correct yaw and roll and the instructor shouting at me to “Follow the bloody flight director needles”. I learned a good lesson from that tirade of abuse on how not to instruct if ever I became a check pilot. In later years, having gravitated to the exalted – or despised maybe – role of simulator instructor, my habit was to introduce the engine failure on take-off by first personally demonstrating to the student how it should be done on raw data; meaning without a flight director. I hoped by first demonstrating, the student could see the body angles or attitude rather than imagine them by trying to “look through” the dancing needles of the FD.
I have always been an advocate of the Central Flying School instructional technique of demonstrate first so the student then knows what is aiming for. Of course in the simulator, the instructor runs the risk of stuffing up (been there - done that!) but it at least proves he is human and not just another screaming skull.
Recently, a 250 hour pilot with a type rating on the 737-300 and trained overseas, booked a practice session prior to putting himself up to renew an instrument rating. His last rating was on a B76 Duchess. As part of the 737 instrument rating would include manual flying on raw data, he was given a practice manual throttle raw data take off and climb to 3000 ft. He protested, saying he had never flown the simulator without the flight director. His instructions were to maintain 180 knots with Flaps 5 on levelling. He was unable to cope and when the instructor froze the simulator to save more embarrassment, the student was 2000 ft above cleared level and 270 knots still accelerating with take off thrust.
The student had been totally reliant on following flight directors with their associated autothrottles during his type rating course and without this aid he was helpless. I believe this is more widespread than most of us would believe; especially as we tend to move in our own narrow circle of experience.
At a US flight safety symposium, a speaker made the point that it is the less experienced first officers starting out at smaller carriers who most need manual flying experience. And, airline training programs are focused on training pilots to fly with the automation, rather than without it. Senior pilots, even if their manual flying skills are rusty, can at least draw on experience flying older generations of less automated planes.
On 4 January this year, the FAA published a Safety Alert for Operators (SAFO) entitled Manual Flight Operations. The purpose of the SAFO was to encourage operators to promote manual flight operations when appropriate. An extract from the SAFO stated that a recent analysis of flight operations data (including normal flight operations, incidents and accidents) identified an increase in manual handling errors and “the FAA believes maintaining and improving the knowledge and skills for manual flight operations is necessary for safe flight operations. ”
Now let me see, I recall similar sentiments nearly 50 years ago published in “Handling the Big Jets” when David Davies wrote ” that airline pilots should “not become lazy in your professional lives…the autopilot is a great comfort, so is the flight director and approach coupler but do not get into the position where you need these devices to complete the flight.” See my earlier paragraphs.
It is a good bet that lip service will be paid by most US operators to the FAA recommendation to do more hand flying. It may have some effect in USA but certainly the majority of the world’s airlines, if they were even aware of the FAA stance in the first place (very doubtful), will continue to stick with accent on full automation from lift off to near touch-down and either ban or discourage their pilots from hand flying on line. If you don’t believe that, consider the statement in one European 737 FCOM from 20 years ago that said: “Under only exceptional circumstances will manual flight be permitted”
After all, when at least two major airlines in SE Asia have recently banned all take off and landings by first officers because of their poor flying ability, then what hope is there to allow these pilots to actually touch the controls and hand fly in good weather? One of those airlines requires the first officer to have a minimum of five years on type before being allowed to take off or land while the other stipulates the captain will do all the flying below 5000 ft.
It might stop QAR pings and the captain wearing the consequences of the first officer’s lack of handling ability, but it sure fails to address the real cause and that is lack of proper training before first officers are shoved out on line.
I think the FAA missed a golden opportunity in its SAFO to note that practicing hand flying to maintain flying skills will better attain that objective if flight director guidance is switched off. The very design of flight director systems concentrates all information into two needles (or V-bar) and in order to get those needles centred over the little square box, it needs intense concentration by the pilot.
Normal instrument flight scan technique is degraded or disappears with the pilot sometimes oblivious to the other instruments because of the need to focus exclusively on the FD needles. Believe me we see this in the simulator time and again. Manual flying without first switching off FD information will not increase basic handling or instrument flying skills.
The flight director is amazingly accurate provided the information sent to it is correct. But you don’t need it for all stages of flight. Given wrong information and followed blindly, it becomes a fatal attraction.
Yet we have seen in the simulator a marked reluctance for pilots to switch it off when it no longer gives useful information. Instructors are quick to blame the hapless student for not following the FD needles. This only serves to reinforce addiction to the FD needles as they must be right because the instructor keeps on telling them so. For type rating training on new pilots, repeated circuits and landings sharpen handling skills. Yet it not uncommon for instructors to teach students to enter waypoints around the circuit and then exhort the pilots “fly the flight director” instead of having then look outside at the runway to judge how things are going.
First officers are a captive audience to a captain’s whims. If the captain is nervous about letting his first officer turn off the flight director for simple climbs or descents, or even a non-threatening instrument approach, then it reflects adversely on the captain’s own confidence he could handle a non-flight director approach. The FAA has already acted belatedly in publicly recommending that operators should encourage more hand flying if conditions are appropriate. But switch off the flight directors if you want real value for money particularly with low hour pilots. It may save lives on the proverbial dark and stormy night and the generators play up.
My first encounter with flight directors was in 1966 while undergoing conversion to the Avro 748. The RAAF had seen fit to send me to Woodford in Cheshire, all the way from Australia to ferry the second of several new 748’s for the RAAF VIP squadron at Canberra. The conversion was conducted on a battered 748 demonstrator G-ARAY known as Gary. The contract allowed four hours of dual for the captains and nothing for the co-pilots.
G-ARAY had the basic instrument flying panel of that era and no flight director. Our instructors at Avro’s were well known test pilots Bill Else, Tony Blackman and Eric Franklin. Jimmy Harrison was chief test pilot. Unlike the bog-standard civilian 748, the RAAF 748’s were to be equipped with a Collins FD 108 FD. So the situation existed that the RAAF 748’s had a British Smith’s autopilot system which was married (somewhat expensively and painfully) to the American Collins FD 108. For the life of me I could not see why a flight director was needed in the RAAF 748. After all, the approach speed was that of a DC3 – 80 knots and the aircraft a delight to handle compared with the venerable Dak.
In retrospect, I think the old Wing Commander Transport Ops at Department of Air, who was charged with the procurement of the 748 for RAAF service, and hadn’t flown for years, was perhaps conned by the Avro sales people in conjunction with Collins, into buying the Collins systems. Certainly in my view as the squadron QFI, flight directors were not operationally needed. In the event, the RAAF machines came with Collins FD 108 flight directors and as the contract specified each captain would be given only one hour of dual instruction once the 748 came out of the factory, we needed to learn how to operate the FD.
First, a course was arranged at the Collins establishment at Weybridge in Surrey. The two RAAF captains and their co-pilots attended and our two navigators and our instrument fitters also turned up to enjoy the Collins hospitality. We learned about 45 degree automatic intercepts of the VOR and ILS beams and other goodies including V-bar interpretation. We were showered with glossy brochures of the flight director by white dust-coated lecturers and shown a film.
By lunch time the presentation was complete and we were shouted to a slap up pub meal with lots of grog, all paid for by Collins. We asked what further lectures were to take place after lunch. We were told the course was over – it was just a morning’s job and we were free to leave unless we would like more drinks. Naturally it was churlish to refuse and hours later we staggered to the railway station (I think), smashed to the eye balls and having forgotten all about the marvels of 45 degree auto intercepts on the FD 108. I must say it was a bloody good three hour course what with the free grog and all that.
A few weeks later, I flew the second RAAF aircraft out of the factory, A10-596, under the watchful eye of Eric Franklin DFC and he demonstrated flight director stuff. For example to climb using the FD, you first put the aircraft into a normal climb and when settled you switched on the FD and carefully wound up the pitch knob so that the little aeroplane sat in the middle of the V-bars.
I quickly realised that you hand flew the basic artificial horizon to whatever attitude was appropriate for the manoeuvre then told the FD 108 V bars where you wanted them. The ILS intercept of 45 degrees was never used because radar vectors didn’t do such angles. I became more and more convinced the 748 didn’t need flight directors and that they were a load of bollocks in that type of low speed aircraft. We were told the USAF used the FD 108 in its F4 Phantoms and that Collins was anxious to makes sales in the UK market. The RAAF Wing Commander got sucked in by good sales talk and from then on all RAAF 748’s became so equipped.
I held personal doubts about the usefulness of flight directors in general as I could see even then, their extended use could lead to degradation of pure instrument flying skills. Today’s flight director systems are light years ahead in sophistication compared with the old Collins FD 105 and 108 series. But the problem with blind reliance on FD indications and thus steady degradation of manual instrument flying skills is as real now as it was back in 1966.
Now to the present day - although first some background history. First published in 1967, “Handling the Big Jets” written by the then British Air Registration Board’s chief test pilot David Davies, is still considered by some as the finest treatise still around on jet transport handling. Indeed, the book was described by IFALPA as `the best of its kind in the world`…written by a test pilot for airline pilots, the book is likely to become a standard text book…particularly recommended to all airline pilots who fly jets in the future…valuable to those pilots who are active in air safety work”.
All that was back in 1967 and little has changed since then - apart from an increasing propensity for crashes involving loss of control rather than simply running into hills. LOC instead of CFIT. Mostly these accidents were caused primarily by poor hand flying and instrument flying skills which certainly explains why aircraft manufacturers lead the push for more and more automatics. A colleague involved with Boeing 787 training was told by a test pilot on type, that the 787 design philosophy was based on the premise that incompetent crews would be flying the aircraft and that its sophisticated automatic protection systems were in place to defend against incompetent handling. Be it a tongue-in-cheek observation, it contains an element of truth.
With the plethora of inexperienced low hour cadet pilots going directly into the second in command seats in many airlines in Asia, the Middle East and Europe, these protection systems are important.
Towards the end of his book, David Davies discusses the limitations of the flight instruments in turbulence and in particular the generally small size of the active part of the basic attitude information or the ‘little aeroplane” as many older pilots will remember it. He continues: “the preponderance of flight director and other information suppresses the attitude information and makes it difficult to get at” and “the inability, where pitch and roll information is split, to convey true attitude information at large pitch and roll angles in combination” Finally Davies exhorts airline pilots “not to become lazy in your professional lives…the autopilot is a great comfort, so is the flight director and approach coupler…but do not get into the position where you need these devices to complete a flight”. There is more but go and read the book.
Having done the unforgiveable and quoted freely from an eminent authority, it is time for Centaurus to say something original and accept the no doubt critical comment that is freely available on Pprune. Flight Directors can be a fatal attraction to those pilots who have been brain-washed by their training system to rely on them at all times. While Boeing in their FCTM advise pilots to ensure flight director modes are selected for the desired manoeuvre, it also makes the point that the FD should be turned off, if commands are not to be followed.
Recently a new pilot to the Boeing 737 asked his line training captain if he could turn off the FD during a visual climb so he could better “see” the climb attitude. His request was refused as being “unsafe” and instead he was told to “look through” the FD. I don’t know about you, but I find it impossible to “see” the little aeroplane when it is obscured by twin needles or V-bars. In fact it takes a fair amount of imagination and concentration to do so. Which may be why Boeing recommends pilots to switch off the FD if commands are not to be followed.
I well recall my first simulator experience in the 737 of an engine failure at V2 where I was having a devil of a time trying to correct yaw and roll and the instructor shouting at me to “Follow the bloody flight director needles”. I learned a good lesson from that tirade of abuse on how not to instruct if ever I became a check pilot. In later years, having gravitated to the exalted – or despised maybe – role of simulator instructor, my habit was to introduce the engine failure on take-off by first personally demonstrating to the student how it should be done on raw data; meaning without a flight director. I hoped by first demonstrating, the student could see the body angles or attitude rather than imagine them by trying to “look through” the dancing needles of the FD.
I have always been an advocate of the Central Flying School instructional technique of demonstrate first so the student then knows what is aiming for. Of course in the simulator, the instructor runs the risk of stuffing up (been there - done that!) but it at least proves he is human and not just another screaming skull.
Recently, a 250 hour pilot with a type rating on the 737-300 and trained overseas, booked a practice session prior to putting himself up to renew an instrument rating. His last rating was on a B76 Duchess. As part of the 737 instrument rating would include manual flying on raw data, he was given a practice manual throttle raw data take off and climb to 3000 ft. He protested, saying he had never flown the simulator without the flight director. His instructions were to maintain 180 knots with Flaps 5 on levelling. He was unable to cope and when the instructor froze the simulator to save more embarrassment, the student was 2000 ft above cleared level and 270 knots still accelerating with take off thrust.
The student had been totally reliant on following flight directors with their associated autothrottles during his type rating course and without this aid he was helpless. I believe this is more widespread than most of us would believe; especially as we tend to move in our own narrow circle of experience.
At a US flight safety symposium, a speaker made the point that it is the less experienced first officers starting out at smaller carriers who most need manual flying experience. And, airline training programs are focused on training pilots to fly with the automation, rather than without it. Senior pilots, even if their manual flying skills are rusty, can at least draw on experience flying older generations of less automated planes.
On 4 January this year, the FAA published a Safety Alert for Operators (SAFO) entitled Manual Flight Operations. The purpose of the SAFO was to encourage operators to promote manual flight operations when appropriate. An extract from the SAFO stated that a recent analysis of flight operations data (including normal flight operations, incidents and accidents) identified an increase in manual handling errors and “the FAA believes maintaining and improving the knowledge and skills for manual flight operations is necessary for safe flight operations. ”
Now let me see, I recall similar sentiments nearly 50 years ago published in “Handling the Big Jets” when David Davies wrote ” that airline pilots should “not become lazy in your professional lives…the autopilot is a great comfort, so is the flight director and approach coupler but do not get into the position where you need these devices to complete the flight.” See my earlier paragraphs.
It is a good bet that lip service will be paid by most US operators to the FAA recommendation to do more hand flying. It may have some effect in USA but certainly the majority of the world’s airlines, if they were even aware of the FAA stance in the first place (very doubtful), will continue to stick with accent on full automation from lift off to near touch-down and either ban or discourage their pilots from hand flying on line. If you don’t believe that, consider the statement in one European 737 FCOM from 20 years ago that said: “Under only exceptional circumstances will manual flight be permitted”
After all, when at least two major airlines in SE Asia have recently banned all take off and landings by first officers because of their poor flying ability, then what hope is there to allow these pilots to actually touch the controls and hand fly in good weather? One of those airlines requires the first officer to have a minimum of five years on type before being allowed to take off or land while the other stipulates the captain will do all the flying below 5000 ft.
It might stop QAR pings and the captain wearing the consequences of the first officer’s lack of handling ability, but it sure fails to address the real cause and that is lack of proper training before first officers are shoved out on line.
I think the FAA missed a golden opportunity in its SAFO to note that practicing hand flying to maintain flying skills will better attain that objective if flight director guidance is switched off. The very design of flight director systems concentrates all information into two needles (or V-bar) and in order to get those needles centred over the little square box, it needs intense concentration by the pilot.
Normal instrument flight scan technique is degraded or disappears with the pilot sometimes oblivious to the other instruments because of the need to focus exclusively on the FD needles. Believe me we see this in the simulator time and again. Manual flying without first switching off FD information will not increase basic handling or instrument flying skills.
The flight director is amazingly accurate provided the information sent to it is correct. But you don’t need it for all stages of flight. Given wrong information and followed blindly, it becomes a fatal attraction.
Yet we have seen in the simulator a marked reluctance for pilots to switch it off when it no longer gives useful information. Instructors are quick to blame the hapless student for not following the FD needles. This only serves to reinforce addiction to the FD needles as they must be right because the instructor keeps on telling them so. For type rating training on new pilots, repeated circuits and landings sharpen handling skills. Yet it not uncommon for instructors to teach students to enter waypoints around the circuit and then exhort the pilots “fly the flight director” instead of having then look outside at the runway to judge how things are going.
First officers are a captive audience to a captain’s whims. If the captain is nervous about letting his first officer turn off the flight director for simple climbs or descents, or even a non-threatening instrument approach, then it reflects adversely on the captain’s own confidence he could handle a non-flight director approach. The FAA has already acted belatedly in publicly recommending that operators should encourage more hand flying if conditions are appropriate. But switch off the flight directors if you want real value for money particularly with low hour pilots. It may save lives on the proverbial dark and stormy night and the generators play up.
Moderator
Thread Starter
Thank you Centaurus, that was excellent!
I have some times flown aircraft equipped with flight directors, but I have never used one. The pilots who mentored my instrument flying spent more time covering some up, or turning them off, rather than indulging me in more technology.
We spend a lot of certification effort standardizing cockpits - instrument position, and appearance, standardizing control location, motion, effect and knob shape, all to reduce workload, and possible task saturation. Then, we introduce a whole suit of new systems, which the marketing guys will tell you makes the plane more foolproof, and easy to fly, but really takes away from the standardization, and introduces confusion to all but the most familiar pilot.
Even in the most early days of VFR GPS, the only buttons you could depend upon for having a common function to all other brands of GPS were the "direct" D with arrow through, and "enter". Seemingly, all other buttons were different brand for brand. When I used to ferry lots of fancy GA aircraft, you'd get one comm working for you, and spend the first hour or two of the flight figuring out everything else!
The tech guys will design more fancy stuff, and tell us how safe it will make flying, the owner want's the flashiest plane and panel, so he'll buy and install it. The pilot wants to push the buttons on the new box, so the basics of flying the plane fall by the wayside - and that's what must always be the most safe, instinctive skill!
I have some times flown aircraft equipped with flight directors, but I have never used one. The pilots who mentored my instrument flying spent more time covering some up, or turning them off, rather than indulging me in more technology.
We spend a lot of certification effort standardizing cockpits - instrument position, and appearance, standardizing control location, motion, effect and knob shape, all to reduce workload, and possible task saturation. Then, we introduce a whole suit of new systems, which the marketing guys will tell you makes the plane more foolproof, and easy to fly, but really takes away from the standardization, and introduces confusion to all but the most familiar pilot.
Even in the most early days of VFR GPS, the only buttons you could depend upon for having a common function to all other brands of GPS were the "direct" D with arrow through, and "enter". Seemingly, all other buttons were different brand for brand. When I used to ferry lots of fancy GA aircraft, you'd get one comm working for you, and spend the first hour or two of the flight figuring out everything else!
The tech guys will design more fancy stuff, and tell us how safe it will make flying, the owner want's the flashiest plane and panel, so he'll buy and install it. The pilot wants to push the buttons on the new box, so the basics of flying the plane fall by the wayside - and that's what must always be the most safe, instinctive skill!
