::  Site Map  ::
			Home  |  About This Website  |  Contact Us

Home » All Evidence from Observational Studies All Evidence from Observational Studies 47 matching pieces of evidence found.   Evidence Type: Excerpt from Observational Study Evidence: "As has been noted, today's tightly coupled automation systems have become extremely complex and in many cases, relatively opaque to their operators. At the same time, these systems have limits which may or may not be clear to their operators. An example of the problems that can be created is seen in this information, extracted form a 1991 incident report: 'Flight XXX departed on schedule; heavy rain and gusty winds were experienced on takeoff and during the departure. The climbout was normal until approximately FL 240 when numerous caution/warning messages began to appear, indicating a deteriorating mechanical condition. The first ... was OVHT ENG 1 NAC, closely followed by BLEED DUCT LEAK L, ENG 1 OIL PRESSURE, FLAPS PRIMARY, FMC L, STARTER CUT OUT 1, and others. #1 generator tripped off line and the #1 engine amber "REV" indication appeared. However, no yaw control problems were noted. The maximum and minimum speed references on the airspeed (tape) came together, followed by stick shaker activation. At approximately FL 260, the cabin was climbing rapidly and could not be controlled. The Captain initiated an emergency descent and turnback to the departure airport. The crew began to perform emergency procedures and declared an emergency. During the descent, the stick shaker activated several times but ceased below FL 200. Due to the abnormal flap indication and the #1 engine reverse, airspeed during the descent was limited to 260-270 knots. The Captain called upon the two augmented crew pilots to assst during the remainder of the flight. While maintaining control of the aircraft, he directed the first officer to handle ATC communications and to accomplish multiple abnomal procedures with the help of the additional first officer. The additional captain maintained communications with the lead flight attendant and company operations as the emergency progressed and later assisted in the passenger evacuation. Fuel dumping began on descent below 10,000 feet. The fuel jettison procedure was complicated as the left dump nozzle appeared inoperative. The crew dumped 160,000 lb of fuel; this action took about 40 minutes. When the fuel dumping was completed, the captain requested vectors for a 20 mile final for runway XX. The crew extended flaps early using alternate procedures due to an abnormal leading edge indication and the FLAPS PRIMARY message ... A final approach speed of Vref + 20 and 25 [degrees] of trailing edge flaps was planned. They selected auto brakes number 4. The weather was still bad with strong, gusty winds and heavy rain causing moderate turbulence during the approach. The ILS approach and landing were normal. At touchdown, maximum reverse was selected on #2 and #3 engines and about half reverse on #4 engine... As the aircraft passed a taxiway turnoff, the tower advised that they saw fire on the left side of the aircraft... This was the first time crew members were aware of any fire... A runway turnoff was used, and the aircraft stopped on a taxiway ... (a difficult but successful evacuation followed). This incident is an example of an electronic system 'nightmare'. The crew received and had to sort out 42 EICAS messages, 12 caution/warning indications, repeated stick shaker activation and abnormal speed reference information on the promary flight display. Many of these indications were conflicting, leading the crew to suspect number one engine problems when that engine was actually functioning normally. There was no indication of fire presented to the crew a fire actually existed...' " (page 155-156) Issue: automation may be too complex (Issue #40) See Issue details Strength: +1 Aircraft: unspecified Equipment: automation Source: Billings, C.E. (1996). Human-Centered Aviation Automation: Principles and Guidelines. NASA Technical Memorandum 110381. National Aeronautics and Space Administration. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: A widely reported problem in modern aircraft is entering instructions through the keypad into the FMS, when under time pressure (e.g. [12]). Pilots have mentioned this issue to us as a particular problem: “...during the high workload phases, operating the FMS, especially the tasks that you don’t do very often, and therefore you might forget to put a slash or a stroke, whatever the format should be that you are typing into the scratchpad, that is very distracting; getting the format correct, especially the format that you don’t often use”. (page 5) Issue: automation may adversely affect pilot workload (Issue #79) See Issue details Strength: +1 Aircraft: unspecified Equipment: automation & FMS Source: Bruseberg, A., & Johnson, P. (2004). Considering temporal aspects for the design of humancomputer collaboration: identifying suitable foci. Department of Computer Science, University of Bath. Available at http://www.cs.bath.ac.uk/~anneb/chi%20time%20ws%202004.pdf. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: A widely reported problem in modern aircraft is entering instructions through the keypad into the FMS, when under time pressure (e.g. [12]). Pilots have mentioned this issue to us as a particular problem: “...during the high workload phases, operating the FMS, especially the tasks that you don’t do very often, and therefore you might forget to put a slash or a stroke, whatever the format should be that you are typing into the scratchpad, that is very distracting; getting the format correct, especially the format that you don’t often use”. (page 5) Issue: data entry and programming may be difficult and time consuming (Issue #112) See Issue details Strength: +1 Aircraft: unspecified Equipment: automation & FMS Source: Bruseberg, A., & Johnson, P. (2004). Considering temporal aspects for the design of humancomputer collaboration: identifying suitable foci. Department of Computer Science, University of Bath. Available at http://www.cs.bath.ac.uk/~anneb/chi%20time%20ws%202004.pdf. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: The PF performed only one housekeeping activity: manipulating the controls of the aircraft systems except for cabin temperature. The PNF performed three activities—manipulating the frequency selectors on the communication radios, manipulating the controls on the communication selector panel, and manipulating the cabin temperature controls—that may be classified as housekeeping activities. Of these four housekeeping activities, only manipulating the cabin temperature controls showed a significant effect of level of cockpit automation (F (3, 188) = 4.02,p< .01). For this activity, only the post hoc comparison of the SP-77 versus the 300e was significant (p <.01). The mean frequencies for each aircraft are given in Table 6. (page 15) Issue: automation may adversely affect pilot workload (Issue #79) See Issue details Strength: +5 Aircraft: B737 Equipment: automation Source: Damos, D.L., John, R.S., & Lyall, E.A. (2005). Pilot Activities and the Level of Cockpit Automation. International Journal of Aviation Psychology, 15(3), 251-268. Lawrence Erlbaum Associates, Inc. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: "The commanded-roll failure emulated an AP-commanded roll that exceeded the target bank angle. Analyses for both roll malfunctions and the soft-pitch malfunction are based on time from initial failure to disconnect of the AP by any means (yoke-mounted disconnect, panel disengage, circuit breaker). Times ranged from 1.8 sec to 107.1 sec (means, medians, and ranges are summarized in Table 1). However, 69% of the pilots disconnected within 13 sec of the initial failure and half within 8 sec. These “immediate” disconnects by 18 of the 29 pilots [62%] were defined by sequences in which no other significant actions occurred between failure onset and AP disconnect…Using an RT of 8.7 sec or less as a cutoff value, 93.7% [18 out of 29 pilots or 62 % ] of the sample of immediate responders were included. Eleven pilots [11 out of 29 or 37%] initially chose to manually override the AP prior to their disconnecting the AP, whether by using the control-wheel steering option or by ovirpowering the aileron servo. One extreme outlier was removed, however, reducing the number to 10 [10 out of 29 or 34%] for the examined distribution." (page 160) Issue: failure assessment may be difficult (Issue #25) See Issue details Strength: +1 Aircraft: unspecified Equipment: autoflight: autopilot Source: Damos, D.L., John, R.S., & Lyall, E.A. (1999). Changes in pilot activities with increasing automation. In R.S. Jensen, B. Cox, J.D. Callister, & R. Lavis (Eds.), Proceedings of the 10th International Symposium on Aviation Psychology, 810-815. Columbus, OH: The Ohio State University. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: "Soft roll (roll sensor). The soft roll failure was rated as third in difficulty to diagnose but was rated easiest to correct (by 13 of 26 pilots [50%]). Following removal of one outlier (194 sec) pilot performance was again categorized as immediate disconnect (16 out of 28 or 57%) or manual override (12 out of 28 or 42%). Those categorized as immediate disconnect responses averaged 1 1 .72 sec (range = 4.52-I 6.69), whereas those categorized as manual overrides averaged 37.45 sec after one outlier was removed (range = 13.16-85.14; outlier shown in Figure 2)." (page 161) Issue: failure assessment may be difficult (Issue #25) See Issue details Strength: +1 Aircraft: unspecified Equipment: autoflight: autopilot Source: Damos, D.L., John, R.S., & Lyall, E.A. (1999). Changes in pilot activities with increasing automation. In R.S. Jensen, B. Cox, J.D. Callister, & R. Lavis (Eds.), Proceedings of the 10th International Symposium on Aviation Psychology, 810-815. Columbus, OH: The Ohio State University. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: "Soft pitch (pitch sensor). The soft pitch failure was rated as most difficult to diagnose (by 12 of 26 pilots [46%]) and was rated third easiest to correct, missing a tie fot second by one tally. Performances were again categorized as either immediate disconnect (12 out [of 29 or 41%]) or manual override (17 [out of 29 or 58%]), ... Three pilots never diagnosed the failures [3 out of 29 or 10%,] manually flying the airplane without disconnecting the AP; their scores and one other outlier were removed, leaving 13. Immediate disconnects averaged 17.7 sec (range = 6.5-3 1 .5), and the 13 remaining manual overrides averaged 46.19 (range = 15.2-76.2)." (page 162) Issue: failure assessment may be difficult (Issue #25) See Issue details Strength: +1 Aircraft: unspecified Equipment: autoflight: autopilot Source: Damos, D.L., John, R.S., & Lyall, E.A. (1999). Changes in pilot activities with increasing automation. In R.S. Jensen, B. Cox, J.D. Callister, & R. Lavis (Eds.), Proceedings of the 10th International Symposium on Aviation Psychology, 810-815. Columbus, OH: The Ohio State University. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: "Runaway pitch trim. This failure was different from the others in that only by pulling the pitch trim circuit breaker would the problem be corrected. The interim solution was the AP disconnect/trim interrupt switch. Only three pilots chose the optimal response, depressing and holding the disconnect, then pulling the circuit breaker. Four others depressed and held the disconnect at various times during the recovery. The vast majority of initial responses were yoke AP disconnect (15), followed in frequency by panel-mounted AP-engage switch (5), mode manipulation (2). manual override (2), and pitch trim circuit breaker (1). Data from 4 participants were removed from consideration due to circumstances that contaminated these data. Of the 25 remaining, 21 of the pilots were classified as immediate responders, 2 were classified as manual overriders, and 2 as mode changers. It should also be noted that two pilots never heard the warning tone, possibly due to high-frequency hearing loss, responding only to aircraft performance changes." (page 163) Issue: failure assessment may be difficult (Issue #25) See Issue details Strength: +1 Aircraft: unspecified Equipment: autoflight: autopilot Source: Damos, D.L., John, R.S., & Lyall, E.A. (1999). Changes in pilot activities with increasing automation. In R.S. Jensen, B. Cox, J.D. Callister, & R. Lavis (Eds.), Proceedings of the 10th International Symposium on Aviation Psychology, 810-815. Columbus, OH: The Ohio State University. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: "The commanded-roll failure emulated an AP-commanded roll that exceeded the target bank angle. Analyses for both roll malfunctions and the soft-pitch malfunction are based on time from initial failure to disconnect of the AP by any means (yoke-mounted disconnect, panel disengage, circuit breaker). Times ranged from 1.8 sec to 107.1 sec (means, medians, and ranges are summarized in Table 1). However, 69% of the pilots disconnected within 13 sec of the initial failure and half within 8 sec. These “immediate” disconnects by 18 of the 29 pilots [62%] were defined by sequences in which no other significant actions occurred between failure onset and AP disconnect." (page 160) Issue: failure assessment may be difficult (Issue #25) See Issue details Strength: -1 Aircraft: unspecified Equipment: autoflight: autopilot Source: Damos, D.L., John, R.S., & Lyall, E.A. (1999). Changes in pilot activities with increasing automation. In R.S. Jensen, B. Cox, J.D. Callister, & R. Lavis (Eds.), Proceedings of the 10th International Symposium on Aviation Psychology, 810-815. Columbus, OH: The Ohio State University. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: "Soft roll (roll sensor). The soft roll failure was rated as third in difficulty to diagnose but was rated easiest to correct (by 13 of 26 pilots [50%]). Following removal of one outlier (194 sec) pilot performance was again categorized as immediate disconnect (16 out of 28 or 57%) or manual override (12 out of 28 or 42%). Those categorized as immediate disconnect responses averaged 1 1 .72 sec (range = 4.52-I 6.69), whereas those categorized as manual overrides averaged 37.45 sec after one outlier was removed (range = 13.16-85.14; outlier shown in Figure 2)." (page 161) Issue: failure assessment may be difficult (Issue #25) See Issue details Strength: -1 Aircraft: unspecified Equipment: autoflight: autopilot Source: Damos, D.L., John, R.S., & Lyall, E.A. (1999). Changes in pilot activities with increasing automation. In R.S. Jensen, B. Cox, J.D. Callister, & R. Lavis (Eds.), Proceedings of the 10th International Symposium on Aviation Psychology, 810-815. Columbus, OH: The Ohio State University. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: "Soft pitch (pitch sensor). The soft pitch failure was rated as most difficult to diagnose (by 12 of 26 pilots [46%]) and was rated third easiest to correct, missing a tie fot second by one tally. Performances were again categorized as either immediate disconnect (12 out [of 29 or 41%]) or manual override (17 [out of 29 or 58%]), ... Three pilots never diagnosed the failures [3 out of 29 or 10%,] manually flying the airplane without disconnecting the AP; their scores and one other outlier were removed, leaving 13. Immediate disconnects averaged 17.7 sec (range = 6.5-3 1 .5), and the 13 remaining manual overrides averaged 46.19 (range = 15.2-76.2)." (page 163) Issue: failure assessment may be difficult (Issue #25) See Issue details Strength: -1 Aircraft: unspecified Equipment: autoflight: autopilot Source: Damos, D.L., John, R.S., & Lyall, E.A. (1999). Changes in pilot activities with increasing automation. In R.S. Jensen, B. Cox, J.D. Callister, & R. Lavis (Eds.), Proceedings of the 10th International Symposium on Aviation Psychology, 810-815. Columbus, OH: The Ohio State University. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: "Runaway pitch trim. This failure was different from the others in that only by pulling the pitch trim circuit breaker would the problem be corrected. The interim solution was the AP disconnect/trim interrupt switch. Only three pilots chose the optimal response, depressing and holding the disconnect, then pulling the circuit breaker. Four others depressed and held the disconnect at various times during the recovery. The vast majority of initial responses were yoke AP disconnect (15), followed in frequency by panel-mounted AP-engage switch (5), mode manipulation (2). manual override (2), and pitch trim circuit breaker (1). Data from 4 participants were removed from consideration due to circumstances that contaminated these data. Of the 25 remaining, 21 of the pilots were classified as immediate responders, 2 were classified as manual overriders, and 2 as mode changers. It should also be noted that two pilots never heard the warning tone, possibly due to high-frequency hearing loss, responding only to aircraft performance changes." (page 163) Issue: failure assessment may be difficult (Issue #25) See Issue details Strength: -1 Aircraft: unspecified Equipment: autoflight: autopilot Source: Damos, D.L., John, R.S., & Lyall, E.A. (1999). Changes in pilot activities with increasing automation. In R.S. Jensen, B. Cox, J.D. Callister, & R. Lavis (Eds.), Proceedings of the 10th International Symposium on Aviation Psychology, 810-815. Columbus, OH: The Ohio State University. See Resource details   Evidence Type: Excerpt from Observational Study Evidence: The remaining two communication activities were concerned with conversations between the two pilots. One represented task-relevant conversation; the other, non-relevant conversation. The mean frequencies are shown in Table 5. A two-way (level of cockpit automation, pilot role) ANACOVA performed on task-relevant conversation showed a main effect of level of cockpit automation (F (3, 188) = 4.71; p <.01). The post hoc comparisons of the SP-77 versus the 300a and of the SP-177 versus the 300a were both significant. Of the variables that we expected to be affected by the length of the climbs and descents, non-relevant conversation was the only one that showed a non-significant effect of the covariate (p = .07).