• News
  • Getting control of LOC

Getting control of LOC

Front and centre on all new head-up displays being installed in FedEx Express widebodies will be a single number in a large font that appears only when a pilot needs it most - the g-load on the aircraft.

As an aircraft upset recovery aid, the G-meter's simplicity belies the years of analysis, flight-testing and internal funding that went into FedEx's decision to include it.

The specialised HUD symbology is a product of a broader programme within FedEx starting later this year to arm pilots with the knowledge, simulator skills and equipment to detect and recover large swept-wing aircraft from in-flight upsets - situations where an aircraft for whatever reason is pitched, rolled or yawed beyond the normal limits. Upsets require pilots to assess the unusual attitude and recover the aircraft back to normal parameters or otherwise risk a loss-of-control (LOC) accident.

Flight International joined FedEx experimental test pilot Bob Moreau in a Boeing MD-11 simulator on 18 December at the company's headquarters in Memphis, Tennessee, to perform unusual attitude recoveries with the HUD system and discuss the company's unique approach to upset training, a curriculum that may become a standard training tool for FedEx pilots later this year.

 © FedEx

FedEx is not alone in its efforts as the industry as a whole is mobilising to reduce LOC accidents, events the Flight Safety Foundation says are the top killers of crews and passengers in the airline industry. A list of LOC accidents between 1998 and 2007 compiled by Boeing reveals that there were 22 that resulted in more than 2,051 fatalities worldwide.


The US Federal Aviation Administration, which had largely stayed on the sidelines on LOC prevention despite a long-standing National Transportation Safety Board recommendation to train crews "to respond to sudden, unusual or unexpected aircraft upsets", proposed early in January making certain "special hazard" LOC training a mandatory element of initial and recurrent simulator training for airline pilots.

The NTSB issued its recommendation in 1996 after the fatal 1994 crash near Pittsburgh International airport of USAir Flight 427, a Boeing 737-300 with 132 on board that experienced an upset and subsequent LOC initiated by an uncommanded rudder fault.

Although the industry as a whole came together in the aftermath of Flight 427 to develop the Airplane Upset Recovery Training Aid, a workbook and video explaining how swept-wing aircraft should be recovered from upsets that cause unusual attitudes, accidents continue. Several recent high-profile crashes (see table) that claimed all lives on board reveal that pilots in each case did not properly handle low-speed stall scenarios in the high-altitude phases of flight, a finding safety experts say is a symptom of inadequate knowledge and training in relation to swept-wing jets despite having upset recovery training materials available.


Air transport-certificated pilots do not have "the baseline knowledge that we've believed them to have", says John Cox, chief executive of Washington-based Safety Operating Systems and a former airline pilot and top safety official with the Air Line Pilots Association.

The FAA's new rule, expected to be finalised by late 2010, will give carriers five years after that to implement a host of new training requirements for pilots, including upset-recovery scenarios in the simulator.


Like many airlines, FedEx in the past has voluntarily included some LOC prevention techniques in the classroom and in the simulator using the industry-developed recovery aid, which was recently updated for the second time to include recovering a large swept-wing aircraft in the high-altitude environment.

Unlike the training aid, which assumes the aircraft is controllable through the primary flight controls, FedEx is leaning toward teaching its 4,700 pilots to be more open minded as to the cause of the upset and method of recovery. "We don't assume a normal aircraft," says Moreau. "A big aircraft doesn't get into those situations unless something is amiss."

As such, the FedEx programme is likely to teach pilots to adopt alternate control strategies, using secondary or tertiary flight-control methods, to regain and retain control of the aircraft. FedEx's 4,700 pilots receive five days of flight training a year, composed of one day of ground school, one day of computer-based training and three days of simulator training.

FedEx's assessment in part comes from its own data, including a late 2005 incident where FedEx McDonnell Douglas DC-10 pilots on approach to Minneapolis/St Paul airport experienced an upset caused by the failure of a wing flap mechanism, one of six upsets involving flap or elevator damage over a 20-year span for the carrier.

An internal safety task force that keeps tabs of the carrier's risks decided to launch an advanced manoeuvre - upset recovery training (AM-URT) programme to better understand what types of upsets and advanced recovery techniques could accurately be learned in a simulator, and which might be better suited for a training aircraft. The group also separately recommended that FedEx purchase liquid crystal HUDs and infrared-based enhanced vision systems for its entire fleet by 2015 to boost pilots' situational awareness.

FedEx contracted with aerospace research group Calspan in Buffalo, New York to evaluate the AM-URT with FedEx MD-11 and Airbus A300 line pilots. The test used Calspan's variable stability Bombardier Learjet 25, as well as its aerobatic Beechcraft F33C Bonanza and MD-11 and A300 full-flight simulators.

One key idea that emerged from the two-year study, which involved 20 FedEx line pilots flying in FFS and in aircraft, was that a simulator can deliver positive transfer of learning for unusual attitude recoveries, events the participants had experienced numerous times during all phases of flight training. The study, however, revealed that negative transfer occurred when exposing pilots to upsets (events that lead to an unusual attitude) that must be recovered, a finding that may be linked to inexperience of pilots with upsets.

The industry was reminded of the dangers of "negative" training in simulators in November 2001 when American Airlines Flight 587, an A300, crashed in New York after the first officer over-controlled rudder inputs in response to wake turbulence and snapped the composite vertical stabiliser off the aircraft.

American, as part of its advanced aircraft manoeuvring programme, had been teaching pilots in the simulator to counter such upsets by using rudder control. Unfortunately, the simulators did not adequately portray "the actual large build-up in sideslip angle and sideloads that would accompany such rudder inputs in an actual airplane", says the NTSB final report on the accident.

FedEx's Moreau says even before AA 587, airlines had begun to question how far to push manoeuvre training in simulators, which lack g-cues pilots would experience in an aircraft.


One observed impact caused by the lack of g-cues is that pilots tend to over-control and expedite recoveries in the FFS while doing the opposite in an aircraft, an effect FedEx saw in the Calspan studies and in 30° pitch-up manoeuvres in MD-10 and Boeing 747 aircraft.

By adding the G-meter readout on its HUD, however, researchers found pilots were not only much more aggressive in delivering the correct amount of control input in aircraft, but also slowed their control input forces and timing in the FFS, both desired outcomes.

The finding is aiding FedEx as it develops a long-term strategy for LOC prevention, which will ideally include academic training to teach key concepts, in-flight training to provide "real world relevance" and teach critical recovery skills, and FFS training to teach "key type-specific procedures", says Moreau.

Related Content