Jerry Bresee

Mr. Bresee is a training industry professional with a 40-year career in systems design and development with a focus on aviation and similar high-risk/high-cost tasks and professions. Mr. Bresee supported flight training programs for nine US military platforms from the F-4 to the F-35, led FAA Advanced Qualification Program (AQP) training system design for five certificated US airlines, and contributed to the development of the US Flight Operations Quality Assurance (FOQA) program. 

Beyond Freeze and Reset: Objectively Assessing Crew Interaction Without Interrupting Task Performance

Because a dynamic task performance is destroyed when interrupted, evaluating crew interaction is typically accomplished through observation and Likert-type rating scales, either directly or after the fact using a recording. Such judgements, even when guided by detailed rubrics, are necessarily subjective. Objective, quantitative data collected through bio-signal recording can provide deeper insight into crew interaction, but collecting these data often involves intrusive interfaces that may themselves affect task performance.

One of the few data sources that can be unobtrusively collected is speech. Research into speech characteristics has shown that, regardless of language or dialect, when two people are communicating effectively, the non-verbal aspects of their speech tend to become highly similar. This effect, referred to as prosodic accomodation, is measurable through software tools that process recorded speech. Time-synching these results with video and flight data recordings can provide instructors with unobtrusively collected objective data for measuring crew interaction. Deploying the latest AI capability on large pilot audio datasets can provide further insight into aggregate trends and guide training scenario development.