The Influence of Surprise on Upset Recovery Performance in Airline Pilots

Figures & data

Figure 1. The experimental design.

Figure 2. The stall recovery test scenario and the measured recovery actions.

Figure 3. Time history of the control inputs of Participant 8 (top plot) and the aircraft’s state (bottom plot) in the surprise condition.

Table 1. Description of the four measured performance criteria, with the corresponding Federal Aviation Administration (2015, p. 2) recovery template principles.

Criterion	Corresponding FAA Principle(s)	Requirements to Meet Criterion
C1. Disengage autothrottle first	1	Disengage the autothrottle at least 2.0 s before significant yoke or pedal inputs.
C2. Start with pitch-down control	2a, 3	Give priority to pitch-down control by starting the recovery with pitch-down control inputs. Strong aileron or rudder inputs (> 50% of max) should not occur at around the same moment (within 2.0 s) of starting pitch-down control to meet this criterion.
C3. Sufficient adjustment of loading	2a, 6	Respond (within 2.0 s) to stick-shaker events with significant pitch-down control and maintain significant pitch-down control during stick-shaker activation. Or, apply sufficient pitch-down control to avoid any stick-shaker events. Timing and strength of unloading and reloading should be so that secondary or late stick-shaker events are avoided. Stick-shaker events were defined as secondary if they occurred subsequent to an earlier stick-shaker event, or late if they occurred after the first unloading action; that is, following the first peak of pitch-down control.
C4. Apply pitch-down trim	2b	Using the pitch trim to aid in pitch-down control during the recovery.

Figure 4. V-N diagram of the recovery of Participant 2 in the surprise condition.

Figure 5. Bitmap of the performance criteria of each pilot. PP = participant number; C1 = disengage autothrottle early; C2 = start with pitch-down control; C3 = sufficient adjustment of loading; C4 = use pitch trim.

Table 2. Criteria met in the two conditions.

	Anticipation (Participants Met/Unmet)	Surprise (Participants Met/Unmet)	N	Δ	χ^2	p	Cohen’s d
C1: Disengage autothrottle early	11/9	6/14	20	−5^a	5.10	.024	.69
C2: Start with pitch-down control	16/4	6/14	20	−10^a	13.41	< .001	1.23
C3: Sufficient adjustment of loading	10/10	5/15	20	−5^a	3.94	.047	.61
C4: Use trim	9/11	3/17	20	−6^a	7.07	.008	.85

^a Significant after Holm–Bonferroni correction.

Table 3. The means and standard deviations of the flight parameters in the two conditions.

	Anticipation		Surprise
	M	SD	M	SD	N	Δ	t	p	Cohen’s d
Response time (s)	2.02	.92	4.12	1.08	20	2.10	8.59	< .001	2.09
Recovery duration (s)	22.83	2.43	21.86	2.90	20	−.96	−1.42	.172	.36
Min CAS (kt)	146.9	2.23	144.3	1.16	20	2.63	−5.62	< .001	1.53
Max CAS (kt)	254.5	16.6	260.7	17.7	20	6.22	1.62	.121	.36
Max descent rate (ft/min)	6,502	1,163	7,388	1,309	20	886	2.70	.014	.72
Max Nz (g)	1.50	.09	1.57	.16	20	.07	2.05	.055	.54
Duration 2nd stick-shaker events^a(s)	.99	1.27	1.69	1.94	19	.70	2.03	.057	.43
Altitude loss (ft)	1,508	361	1,693	385	20	186	2.47	.023	.50

Note. CAS = airspeed.

Table 4. The means and standard deviations of the subjective and physiological measures in the two conditions.

	Anticipation		Surprise
	M	SD	M	SD	Δ	n	p	t	Cohen’s d
P. surprise (0–10)	1.39	2.00	8.44	1.50	7.05^a	20	< .001	12.35	3.99
P. startle (0–10)	1.22	2.00	4.28	2.32	3.06^a	20	< .001	5.48	1.41
P. confusion (0–10)	2.78	2.35	3.50	1.92	.72	20	.044	2.16	.34
P. workload (1–20)	12.00	3.18	13.20	3.09	1.20^a	20	.027	2.40	.38
P. anxiety (0–10)	3.28	2.35	4.06	1.81	.78	20	.074	1.89	.37
Δ Mean HR (bpm)	14.69	6.14	14.18	6.40	−.51	15	.805	−.25	.08
GSR (mS)	.05	.10	.19	.31	.14^a	18	.020	2.57	.61

Note. P = perceived; HR = heart rate; bpm = beats per minute; GSR = galvanic skin response.

^a Significant after Holm–Bonferroni correction.

Federal Aviation Administration. (2015). Stall prevention and recovery training (Advisory Circular No. 120-109A). Retrieved from https://www.faa.gov/documentLibrary/media/Advisory_Circular/AC_120-109A.pdf

 Google Scholar