Piloting Skills Undermined by Automation
Pilots are so accustomed to relying on cockpit automation that their basic airmanship skills to fly manually are eroded. The result? When automation fails, pilots cope with the situation either with misguided and distracting attempts to restore the automation or with inappropriate and ham-fisted techniques when suddenly forced to hand-fly the airplane.
The reliance on automation, and some airlines’ encouragement in its use, has led to a situation in which the airplane is probably on autopilot even before the seat-belt sign is turned off after takeoff.
A recent report by the Inspector General of the Department of Transportation (DOT/IG) comes to this glum conclusion:
“FAA [Federal Aviation Administration] does not have a sufficient process to assess a pilot’s ability to monitor flight deck automation systems and manual flying skills, both of which are important for identifying and handling unexpected events.”
The title of the report is revealing: “Enhanced FAA Oversight Could Reduce Hazards Association With Increased Use of Flight Deck Automation” (see https://www.oig.dot.gov/sites/default/files/FAA%20Flight%20Deck%20Automation_Final%20Report%5E1-7-16.pdf). The FAA has overseen, and approved, cockpit automation which has led to a degradation in piloting skills. Now, the same kind of “oversight” is necessary to reverse the trend. Perhaps something more basic is needed.
The trained, professional airman sitting in front may be confused by what his state-of-the-art instruments and displays are, or are not, telling him. Many pilots are so reliant on automation they are reluctant to switch it off, and too many pilots are not confident in their basic hand-flying skills. Recent accidents have proved over-reliance on automation to be a recipe for failure. As demonstrated by the Asiana Airlines B777 accident in 2013 at San Francisco, a pilot may even construct an accident scenario through mishandling the automation — and be quite unaware of what he’s done. (See Aircraft Accident Report, www.ntsb.gov/investigations/AccidentReports/Reports/AAR1401.pdf)
The DOT/IG notes that the FAA has developed new simulator requirements for hand-flying the airplane, which take effect in 2019:
New Simulator Requirements for 2019
|Upset prevention and recovery||Aircraft upset is an unsafe condition which may result in loss of control. Training should focus on the pilot’s manual handling skills to prevent upset, as well as training to recover from this condition.|
|Manually-controlled arrival and departure||Pilots will be both trained and evaluated on their ability to manually fly a departure sequence and arrival into an airport.|
|Slow flight||Pilots will be trained to understand the performance of the aircraft and the way it handles at airspeeds just above the stall warning.|
|Loss of reliable airspeed||Training will focus on the recognition and appropriate response to a system malfunction which results in a loss of reliable airspeed [display] which increases risk of aircraft stall and/or upset.|
|Recovery from stall/stickpusher activation||Training will provide pilots with the knowledge and skills to avoid undesired aircraft conditions which increase the risk of encountering a stall or, if not avoided, to respond correctly and promptly.|
|Recovery from bounced landing||A poorly executed approach and touchdown can generate a shallow bounce (skip) or a high, hard bounce that can quickly develop into a hard landing accident.|
The DOT/IG observed that the FAA is “developing guidance for implementing the new training requirements” but that “a completion date has not been determined.”
The DOT/IG did not indicate that the FAA’s glacial rate of progress is indicative of the low priority accorded the new training protocols.
If pilots cannot confidently and smoothly hand-fly the airplane through all phases of flight, and deal with upsets, unusual attitudes, loss of engine power, loss of electrics/hydraulics/pneumatics, etc. through all phases of flight, the notion of professionalism has been degraded to that of systems monitor. Yet there are situations in which the pilot’s basic skills at the controls will determine the fate of the airplane.
Below is a menu of correctives which will improve pilots’ handling skills:
At No Cost
Every chief pilot and fleet manager should declare the following policies:
— During cruise, aircrew to be strapped in snug, with at least one pilot’s feet within reach of the rudder pedals.
— During approach and landing, pilots strapped in tight, with the flying pilot’s feet resting lightly on the rudder pedals. The feet of the pilot monitoring should be easily within reach of the rudder pedals.
Incorporate the push-power-rudder-roll mantra in ground training for unusual attitude recovery:
— Push the yoke or side stick forward.
— Power, as in increase power to the engines.
— Rudder is applied. The rudder movement induces a yaw. That yawing motion causes the lower wing to move faster than the opposite, upper wing. The downside wing thus generates more lift. To be sure, whenever advocating rudder to be used in this manner, recall the lessons learned from the 2001 crash of American Airlines Flight 587, where the first officer was taught to use the rudder enthusiastically when experiencing attitude excursions. (See Aircraft Accident Report, In-Flight Separation of Vertical Stabilizer, www.ntsb.gov/investigations/AccidentReports/Reports/AAR0404.pdf)
— Roll to wings level.
Include an operating flight strength diagram in aircraft flight manuals, particularly in the “performance limitations” section, so that aircrews better understand the G loadings which result from full deflection of the elevator at a given airspeed.
Aircrews should keep their hands on the flap control lever until deployment is complete (asymmetric deployment can lead to an unusual attitude, which can be more quickly countered if the hand of the pilot flying is still on that secondary control lever).
At Modest Added Cost
Incorporate unusual attitude recovery techniques in initial flight training. Roll angles to 110° can be demonstrated airborne in a light trainer aircraft.
Send a cadre of check pilots and simulator instructor pilots through advanced in-flight unusual attitude recovery training. Thus exposed, they will be loathe to teach pulling the yoke full aft, losing altitude while increasing angle-of-attack, speed and G’s which can exceed the structural limits of the airplane.
Add a real-time G readout at the instructor position in all simulators so that instructors will be better equipped to advise students when they have pulled the wings off of the aircraft.
Perform fully developed stalls in the simulator at low altitude and at cruising altitude to show that with the correct techniques (push-power-rudder-roll) less altitude will be lost at low altitude, and to reinforce that one cannot fly out of an inadvertent stall at high altitude without losing some height.
At Greater Cost
Incorporate an angle of attack (AOA) indicator in the pilot’s primary line of sight, as an integral part of the attitude director indicator/primary flight display (ADI/PFD). Providing AOA information to the pilots provides an added margin of safety. An increasing AOA can warn of icing. Even with a “clean” (uncontaminated by ice) wing, the additional wing loading of a turn at low airspeed can put an airplane into a stall in the blink of an eye. An AOA indicator with a distinctive aural alerter and stick shaker would provide an essential warning to the pilots.
Expand simulator capabilities to present high-fidelity, real-world unusual attitude recovery situations. The goal is to inculcate familiarity, recognition and an instinctive avoidance response.
Ensure adequate tactile feedback, such as moving throttle levers when engine power changes, trim wheels that move when trim changes, and yokes/side sticks with increasing-force feedback to pilot inputs. In the 2009 crash of Air France Flight 447 in the Atlantic, and the 2008 A320 accident off Perpignan, France, the pilots were so unfamiliar with the manual trim wheel that they never even thought about it, let alone used it to extricate themselves from a dire, developing situation. (See Final Report on the Accident on June 1, 2009, to the Airbus A330-203 … Operated by Air France, www.bea.aero/docspa/2009/f-cp090601.en/pdf/f-cp090601.en.pdf. For the Perpignan accident, see Report on Accident off the coast of Canet-Plage at www.bea.aero/docspa/2008/d-la081127.en/pdf/d-la081127.en.pdf)
Train all pilots, in the air, in unusual attitude recovery techniques. Airlines need not maintain expensive trainer aircraft but can, instead, contract out this vital training to certified providers. Actual flight is necessary to reinforce the contrast with simulators. Stall recovery at both low altitude and at cruising altitude must be part of such training to reinforce the contrast. In both the AF 447 and the Perpignan accidents, both airplanes were super-stalled (i.e. they entered a stall condition due to an extreme and unsustainable pilot-induced nose-high pitch attitude selection). AF447 was inadvertently driven nose-high into the so-called “coffin corner” of the flight envelope at altitude The stunned first officer kept his A330’s side-stick aft (sight unseen by his opposite number) in a confused attempt to control the drastic post-stall height-loss In the Perpignan accident, the A320 entered its nose-high extreme stall attitude due to a misguided attempt to mode-change and complete a hectic air-test schedule at a much lower height than was safely appropriate. This distinctive stall onset is very different from any stall entered subtly via inadvertent speed reduction at a virtually constant pitch attitude at lower altitudes. A low level stall onset arrives with very noticeable buffeting and airframe vibration. In a super-stall entered at or near cruising height, the wing’s AOA is so high that turbulent airflows off the wing pass above the tailplane, so there is no attention-getting warning buffet at either the incipient or fully developed stage. The super-stalled airplane has the aerodynamics of a brick, thus the “locked-in” extremely high rates of descent and stable post-stall attitudes seen in those accidents. Moreover, the two airplanes were auto-trimmed into the stall; thus the trimmable horizontal stabilizer was fully deflected and sustained the stall. Manual pitch-trim was available but unused. The underslung engines, at full power, added to this overall pitch-up moment and thus the stalled condition was increasingly stabilized and self-sustained as more height was lost and as thrust increased. All airline pilots should be trained in these inter-relationships and how to recover from a super-stall.
Not only is this menu of actions (and their rationale) more comprehensive than that published by the FAA, it also is far more likely to restore airmanship and professionalism among airline pilots.
In the Asiana accident discussed at length in the DOT/IG report, three “trained” pilots sat on their hands and allowed their jet to slam into a sea-wall well short of the runway on a beautiful July day in 2015. Afterwards, they were quite bereft, bewildered and clueless as to why the accident could have happened. This was a real wake-up call. The profession has been slowly hollowed out by automation and relentless cost cutting. It is time to reflect soberly that the best guarantor of a safe flight is a well-trained and alert pilot skilled in the manipulative nuances of airmanship in extremis.