Loss of Control in Flight: comparing qualitative pilot opinion with quantitative flight data

Loss of Control In Flight (LOC-I) is the most lethal type of accident in recent aviation history and the one that has experienced the least reduction in the past 20 years. These, combined with recent accidents such as Lion Air 610 and Ethiopian Flight 302, help explain the attention dedicated by several aviation stakeholders to prevent it from happening. Prior to that, however, the understanding of the characteristics of the phenomenon is of paramount importance.Pilots, aircraft and systems are deeply related in LOC-I accidents, but current definitions tend to examine these in isolation and not as a human-in-the-loop system. Following experiments in a flight simulator with a group of test pilots, this paper investigates the use of the Cooper-Harper rating scale and the Quantitative Loss of Control Criteria to assess controllability in a series of manoeuvres performed in four different sets of LOC-I test scenarios. The
experimental design, including the selection of pilots, adaptation of test scenarios, conception of tasks and data gathering procedure, is thoroughly discussed. From a statistical correlation analysis, results indicate the existence of a positive weak to moderate monotonic relationship between the qualitative and quantitative approaches, however the resemblance of events indeed
impacting on controllability issues is low. Finally, a discussion is taken on ways to conceive tasks with regard to the application of the Cooper-Harper scale and the possibility of adapting the quantitative method, both in the sense of being more faithful to the actual aircraft condition and pilots’ workload.