Airbus Stick Technique

I normally write about ideas that apply to flying any aircraft. But today is different. This is Airbus specific. Very specific! Recently read a couple of peer-reviewed scientific research papers that do a great job parsing Airbus stick skills and fine-grain aircraft control. I’ve learnt some new things, confirmed some suspicions, but still have questions.

The four topics are:

• Where to hold the side stick?
• Separate up/down and left/right inputs, or move the stick in 2-D?
• Aim for perfectly centered or smoothly tracking?
• What else did the researchers learn?

The main research paper is Stirring the Pot: Comparing Stick Input Patterns and Flight-Path Control Strategies in Airline Pilots by Andreas Haslbeck, Hans-Juergen Hoermann and Patrick Gontar, published in 2018 by the respected International Journal of Aerospace Psychology. Building on earlier research (like The way pilots handle their control stick – effects shown in a flight simulator study, 2012 and Flying the Needles: Flight Deck Automation Erodes Fine-Motor Flying Skills Among Airline Pilots, 2016) lead researcher Andreas Haslbeck, then at Technische Universität München, had 126 fully-qualified European airline pilots flying a manual instrument approach to 26R at Munich in full-motion FFS Level D flight simulators (A320 and A340). Normal law with auto thrust, 1,200 m visibility in light rain, wind 220 at 17 gusting to 22 knots. It was a controlled mix of captains/FOs and short-haul/long-haul experience. The experimental scenario started with a malfunctioning flight director and autopilot 8 min before touchdown. Holding the other factors constant, they recorded stick inputs and resulant flight paths then performed extensive statistical analysis. It’s pretty dang interesting.

Where and how to hold the stick?

No one seems to know! The stick is clearly ergonomically designed for one position, but the researchers found many holding strategies; including high, low, tight, fingertip. In one study about half of pilots applied two or more different grasps within a short period of time. Some pilots grasped the stick constantly for the approach, some opened their hand to relax their muscles several times. There was limited analysis of this data, and I couldn’t see any trends in results.

Me? I simply hold it where I think it was designed, with the same armrest settings every time (G 5, just felt ‘right’ first time in the sim in 2003). I’ve read no flight training guidance. It’s hard for me to see what first officers are doing, and during six months of cockpit safety observations I wasn’t looking at stick grasp position. Where do you hold it? What works best?

Dimensionality of sidestick inputs?

This really is an Airbus only discussion, because of the flight control laws and full-time trim functions. A joystick is obviously a two dimensional input device, and is intuitively used as such on simple conventional aircraft. But in the Airbus stick position is not commanding flight surface position — rather it’s longitudinal g-load (except at very low speed when it’s deck angle change) and lateral roll rate (below 33° bank). Because of this, and their personal experiences, most instructors recommend sequential separate inputs. Roll input, center, pitch input, center. You can sometimes sort of pump stick inputs like a taildragger rudder dance.

The data showed that at higher altitudes on the approach (3,000 to 1,000 feat AGL) the majority of inputs were in a single axis while at low altitudes (between 270 and 50 feet AGL) the majority of inputs were in both axes. This is when we sometimes start ‘stirring the pot’, making more inputs than can be good. We can see the long haul A340 pilots made more two-dimensional inputs.

But what was the result of all these inputs? Analysis clearly showed that:

“Pilots with more one-dimensional inputs (operating roll and pitch axes discretely) achieved overall smaller ILS deviations.”

OK then. This strongly reinforces the guidance given at many airlines — treat stick input axes separately rather than together.

Aim for perfection or smooth steady good enough?

The researchers identified two flight-path control strategies on the ILS: optimizer and steady path. The optimizer makes constant corrections aiming to regain the perfect center of the ILS. The steady path leaves minor deviations uncorrected, resulting in a smoother flight-path. I can see advantages to both ideas. On the localizer “both distinct strategies occurred with approximately equal frequency”, but on the glideslope “the optimizer was clearly the predominant strategy in almost all cases”.

What worked best? The results were clear:

“With one exception (A340, 3,000–1,000 ft AGL, localizer), all optimizer strategies were associated with smaller ILS deviations.”

Interestingly, “the optimizer strategy did not necessarily require more stick deflections” than the smooth steady path idea. I’m going to keep working for nuts-on centered, since other pilots are doing it without more frequent inputs. The steady path strategy may be a comforting story we tell ourselves!

What else?

This is really good research. Lots of expert subjects in a full-motion flight sim doing regular pilot stuff analyzed by PhD-level stats. This is not some kid’s undergrad paper using five private pilot friends in Microsoft Flight Simulator. But still, it has limitations. The differences between long-haul A340 pilots and short-haul A320 pilots are interesting; but could be related to proficiency, size of aircraft, age or experience. It’s hard to tease out the reasons, so more testing with more pilots from different airlines would be fantastic. What the researchers do know is this:

“Short-haul pilots with 16 or 17 landings a month maintain a higher level of fine-motor flying skills than long-haul pilots with only two or three landings a month.” 

Not unexpected. But this is despite the long-haul pilots generally being older and more experienced. Sometimes it’s this simple — if you want to fly better, fly more.

What do you think? How do you hold the Airbus stick? How do you input controls on approach?