Will Controllers Accept a Machine That Thinks like They Think? The Role of Strategic Conformance in Decision Aiding Automation

Abstract

In a series of real time trials, we simulated sophisticated air traffic management conflict resolution automation using unrecognizable replays of controllers’ own performance. Using a fairly novel experimental design and a prototype air traffic control interface, we explored with operational controllers the interactive effects of traffic complexity, level of automation, and “strategic conformance” (defined as the match between human and machine solution strategy) on a number of dependent measures. Conformal advisories (exact replays of a given controller’s previous solution) were accepted more often, rated higher, and responded to faster than were non-conformal advisories (replays of a colleague’s different solution). In the end, one result stood out in particular: roughly 24% of conformal advisories were rejected by controllers. How could it be that controllers, in effect, disagreed with their very own solutions roughly one quarter of the time? The project is currently exploring this and other related issues through extended human-in-the-loop simulations.