Training & recruitment: Raising the bar

Advanced type and recurrent training is so expensive that pilots do not get enough of it to hone all their skills. But does it have to be that way?

Type rating and recurrent airline pilot training has changed little in 30 years. But aircraft and simulator manufacturers are pushing to change that, with advanced fixed-base flight training devices (FTD) and lower-cost full-flight simulators (FFS) among the options.

Quality and quantity of pilot training, particularly at advanced levels, has always suffered because of the cost of using high category "level D" FFSs, or the actual aircraft. Type rating and recurrent training suffer most, and the expense brings enormous pressure to keep costs down. The result is that flightcrew performance suffers because the exercises tend to be limited to those required by law. The same problem can affect ab initio trainees in the final stages of preparation for the line.

A Flight International crew management conference in Brussels this week (23-24 November) will address this need to find a more cost-effective way to provide pilots with advanced training.

Jean-Michel Bigarré, vice-president training and flight operations at regional aircraft manufacturer ATR, is concerned that small carriers wanting to raise their piloting standards cannot do so. "If an airline wants to push its pilots to improve, it faces a finance problem," he says. Convinced that modern simulation could improve training quality while reducing costs, ATR has been working with Canadian simulator manufacturer Mechtronix Systems and the European Joint Aviation Authorities to find a lower-cost recurrent or type training solution acceptable to the regulators. Bigarré was formerly head of training at the French civil aviation authority DGAC.

Bigarré's argument is that if costs could be brought down by reducing the time working pilots spend away from the line, and by using modern FTDs instead of costly FFSs, pilots could go beyond the essential emergency exercises mandated in the regulations. This could provide practice in managing normal situations pilots are likely to encounter, but for which recurrent training is not compulsory. To justify this, Bigarré points to "the increasing numbers of a new kind of accident or incident caused by insufficient [crew] confidence, knowledge and skills". He cites a 9 May 2004 accident in which an American Eagle ATR 72 bounced badly on landing at San Juan, Puerto Rico, wrecking the aircraft.

Gusting crosswind

The US National Transportation Safety Board (NTSB) report says that, on touchdown, the aircraft "skipped once, bounced hard twice, and [finally] came to a complete stop on a grassy area about 217ft [66m] left of the runway 08 centreline". The co-pilot was landing the aircraft in a gusting crosswind and, after the aircraft had bounced heavily, the captain took control, but failed to recover the situation. The NTSB gives the probable cause of the accident as "the captain's failure to execute proper techniques to recover from the bounced landings and his subsequent failure to execute a go-around".

The gusting crosswind conditions on that approach can be programmed into a modern FTD during a recurrent-training operational proficiency check (OPC), says Bigarré - as could many other circumstances likely to be seen in normal flying. But time in an FFS is too expensive to make it practical to replicate such situations. The Flight International conference will examine whether airlines are making a false economy if, instead of using modern simulation technology to extend crew training at low cost, they just pocket savings and train as before.

If a large carrier with a big fleet conducts its own pilot training in-house, it may decide to invest in one or more flight simulators. That way, its pilots do not have to waste time travelling to a third-party training organisation every six months, adding to the time they are absent from the line and running up travel and hotel bills. Smaller carriers, however, cannot justify purchasing their own simulators and have to send their pilots to commercial training organisations.

But if airlines were allowed to carry out all or most of the training on a modern fixed-base (F-B) FTD with a high-quality visual system, the smaller carriers might be able to operate one, or share one with others, obviating the need to send pilots away from base as well as cutting the per-hour cost of training. Motion-base FFS not only cost about five times more than a F-B FTD, according to Bigarré, but also need special high-ceiling buildings and incur considerable maintenance costs.

JAA approval

ATR's Toulouse-based training centre has achieved JAA approval for an F-B FTD to be used for pilot OPCs during recurrent training. This is the first F-B device ever to be certificated for the full OPC, says Bigarré. It provides high-fidelity reproduction of the aircraft's cockpit and controls, and has a collimated (infinity-focused) visual system that provides a degree of realism, plus simulated vibration and ambient noise, that largely compensates for the lack of six-axis motion. Add an optional pilot-seat actuation system to provide "seat-of-the-pants" sensations of turbulence, runway surface roughness or airframe vibration, and Bigarré claims the device has become "a fixed-base FFS".

Mechtronix president Xavier Hervé calls the system an FFT - a full-flight trainer - to distinguish it from traditional fixed-base devices. Using the same simulation software as the full-flight simulator, based on manufacturer-supplied data packages, and with high-fidelity aerodynamic, ground and engine performance and control forces, the FFT provides "FFS quality without motion", says Hervé. ATR Training's 72-500 "FFT" is also cleared for all but the final 16h of a first-time type-rating, he adds.

He says regulators are aware of advances in simulation capability and are prepared to take advantage of the latest training tools. This has led to a rekindling of interest in using a modern Level B full-flight simulator - with visual and motion - as an alternative to the higher-fidelity, but more expensive Level D "zero flight time" FFS for recurrent training. Mechtronix has already sold such devices to Panama's Copa Airlines, Hungary's Malev and Lufthansa. Copa's Boeing 737-800 FFS X, for example, is certificated level B, but Hervé insists: "The key is not the level of qualification by itself, it is the second axis - the level of qualification for training capabilities." The simulator's extra performance includes TCAS [traffic collision avoidance system] alerts and manoeuvring practice, low visibility CAT III approaches, and wind-shear training. Malev's FFS is also level B, but it is approved for all initial and transition type-rating training tasks except the final checks of take-off and landing techniques.

Bigarré questions whether motion is necessary at all - and, where it is deemed necessary, whether the system needs to be as massive as it is. Hervé argues effective motion systems do not need as much actuator travel as believed, and electrically driven systems also bring down the cost.

Hervé wonders whether it is necessary for a type-specific Level D FFS to use actual aircraft instruments and systems. Bigarré agrees, pointing out: "FFS instruments and systems are specific and therefore expensive. New technologies may be used to create realistic and affordable cockpits."

The issue of whether full-motion FFS should be the only acceptable synthetic way of performing certain types of pilot training is under test. The FAA and NASA are working with line pilots "flying" specific "scenarios", some with failures or windshear introduced. They compared pilot training performance using full motion with the outcome when the simulator's motion was switched off. The learning process was quicker, in some cases, when motion was not used, and the pilot's physical reaction times were better.

Sensory systems

Motion is used to fool the pilots' sensory systems - for example, by tipping the simulator backwards so that gravity provides a "push" into the seat-back that simulates forward acceleration - but the flight instruments tell a different story about pitch attitude and this puts the pilots' inner-ear balance sensors at odds with what their eyes tell them. Sustained yaw or sideslip acceleration can be simulated by tilting, but vertical acceleration can be simulated only momentarily. NASA tests bear out what pilots have always known - that they are more likely to over-control in a simulator than in a real aircraft, and they over-control less with the motion off.

Many operators cannot afford simulation, so cannot train their pilots for situations that are too risky to carry out in the real aircraft. Modern fixed-based FTDs and Level B FFSs are potential solutions. Bigarré wants to push the use of simulators so airlines can afford to widen their training beyond the regulatory basics.

Source: Flight International