Single-Pilot Workload Management in Entry-Level Jets

Civil Aerospace Medical Institute2013PDFTech ReportBurian, Barbara K.Pruchnicki, ShawnRogers, JasonChristopher, BonnyWilliams, KevinSilverman, EvanDrechsler, GenaMead, AndyHackworth, CarlaRunnels, BarryAmes Research CenterSan Jose State UniversityCivil Aerospace Medical InstituteUnited States. Office of Aerospace MedicineUnited StatesAir pilotsWorkloadSelf evaluationHuman factorsAutomationJet propelled aircraftCountermeasuresDOT/FAA/AM-13/17Researchers from the NASA Ames Flight Cognition Lab and the FAA\u2019s Flight Deck Human Factors Research Laboratory at the Civil Aerospace Medical Institute (CAMI) examined task and workload management by single pilots in Very Light Jets (VLJs), also called Entry-Level Jets (ELJs). Fourteen certificated Cessna Citation Mustang (C510-S) pilots flew an experimental flight with two legs involving high workload management under Instrument Flight Rules (IFR) in a Cessna Citation Mustang ELJ level 5 flight training device at CAMI. Eight of the pilots were Mustang owner-operators, and the other six flew the Citation Mustang as part of their jobs as professional pilots. In addition to the Cessna Citation Mustang simulator, data collection included instantaneous self-assessment of perceived workload, NASA Task Load Index (TLX) workload measures, researcher observations, final debriefing interviews, and three questionnaires: Cockpit Set-up Preferences, Demographics, and Automation Experiences and Perceptions. To facilitate analysis, the major high workload tasks during the cruise portion of flight were grouped into four events. Approximately two-thirds of the tasks within the four events were accomplished by the participants with no difficulties. Though all participants committed a variety of errors during all four high workload events (e.g., readback error, airspeed violation), most errors were not directly related to overall task success. We did find a significant effect on task performance success related to hours of experience only for the first event. We also found that some type of error using the G1000 avionics was at the root of the problem for most participants who had difficulty accomplishing one or more of the tasks. All participants committed a variety of errors during all four high workload events (e.g., readback error, airspeed violation), but most were not directly related to overall task success. Implications of the findings are discussed, and techniques demonstrated by our participants that we have characterized as \u201cbest practices\u201d have been identified. Recommended strategies for automation use and countermeasures to task overload and workload breakdowns have also been provided.119