As new technology and procedures are envisioned for the future airspace, it is important to predict if these may compromise safety in terms of pilots' failing to notice unexpected events. Computational models such as N-SEEV support cost-effective means of making such predictions.

No abstract

A taxonomy of navigation er r or s (pilot deviations) dur ing taxi oper ations was developed that defines 3 classes of er r or s: planning, decision, and execution err or s. This taxonomy was applied to er r or data fr om 2 full-mission simulation studies (Hooey, Foyle, Andr e, & Par ke, 2000;McCann et al., 1998) that included tr ials that r eplicated cur r ent-day oper ations and tr ials with advanced cockpit technologies including datalink, electr onic moving maps (EMM), and head-up displays (HUDs). Pilots committed navigation er r or s on 17% of cur r ent-day oper ations tr ials (in low-visibility and night), distr ibuted r oughly equally acr oss the 3 er r or classes. Each er r or class was associated with a unique set of contr ibuting factor s and mitigating solutions. Planning er r or s wer e mitigated by technologies that pr ovided an unambiguous r ecor d of the clear ance (datalink and the EMM, which possessed a text-based clear ance). Decision er r or s wer e mitigated by technologies that pr ovided both local and global awar eness including information about the distance to and dir ection of the next tur n, cur r ent ownship location, and a gr aphical depiction of the r oute (as pr ovided by the EMM and HUD together ). Execution er r or s wer e best mitigated by the HUD, which disambiguated the envir onment and depicted the clear ed taxi r oute. Implications for technology design and integr ation ar e pr ovided.

The Man-machine Integration Design and Analysis (MIDAS) human performance model was augmented to improve predictions of multi-operator situation awareness (SA). In MIDAS, the environment is defined by situation elements (SE) that are processed by the modeled operator via a series of sub-models including visual attention, perception, and memory. Collectively, these sub-models represent the situation assessment process and determine which SEs are attended to and comprehended by the modeled operator. SA is computed as a ratio of the Actual SA (the number of SEs that are detected or comprehended) relative to the Optimal SA (those deemed required or desired for the operator to complete his/her task).A high-fidelity application model of a two-pilot commercial crew during the approach phase of flight was generated to demonstrate and verify the SA model. Two flight deck display configurations, hypothesized to support pilot SA at differing levels, were modeled. The results presented include the ratio of actual to optimal SA for three high-level tasks: Aviate, Separate, and Navigate. The model results verified that the SA model operates as expected and is sensitive to scenario characteristics including display configuration and pilot responsibilities.