Insights from a Simulator Study: How the Intelligent Pilot Advisory System (IPAS) Could Impact Decision-Making in Flight Operations

Abstract

The Intelligent Pilot Advisory System (IPAS), developed by the German Aerospace Center (Deutsches Zentrum für Luft- und Raumfahrt (DLR)), is an AI-based cockpit assistance system that supports decision-making in normal, abnormal, and emergency flight situations. Previous work on IPAS has mainly addressed conceptual, technical, or explainability aspects and has not examined its operational impact. It is still unknown how automating parts of the pilots’ decision-making process affects the decision-making structure, crew collaboration, information usage, and how the crews perceive the assistance in terms of usefulness and operational AI explainability. This exploratory study looked at six flight crews performing simulator flights involving technical malfunctions requiring diversions, both with and without the IPAS. The IPAS assists the crew by automatically identifying, assessing and selecting alternative airports based on technical, operational and environmental criteria. The crews’ decision-making processes were evaluated through analysis of decision-making duration, expert reviews conducted by an experienced pilot, and questionnaires on system usefulness and explainability. The results indicated improvements in decision-making and collaboration within the crew when using the IPAS. Expert reviews revealed a slight tendency towards uncritical acceptance of the system’s recommendations, suggesting a potential risk of automation bias. This study provides the first operational insights into the integration of the IPAS into the flight deck and its effect on decision-making. Further research with detailed and objective metrics is needed to comprehensively assess the effect of the IPAS on decision-making.