Reduced taxi delays with intelligent departure metering advisory tool (I-MATE): insights from air traffic controller validation experiments

Abstract

Airport taxi delays significantly impact airlines, passengers, and the environment. Effectively addressing taxi delays necessitates the synchronization of stochastic and uncertain airside operations, encompassing aircraft pushbacks, taxiway movements, and runway take-offs. This paper presents the Intelligent Departure Metering Advisory Tool (I-MATE), an AI-based decision support system that recommends pushback timings to Air Traffic Controllers (ATCOs) to reduce taxi delays while balancing downstream runway throughput. I-MATE is fundamentally different from other AI assistants, like conflict resolution tools, in that its recommendation outcomes unfold over a longer time horizon, posing challenges for ATCO comprehension. We conducted validation experiments to assess the efficacy and acceptability of I-MATE in assisting ATCOs to manage airside traffic. The results reveal a significant reduction in taxi delays (25.6%) with increased compliance with I-MATE recommendations, which may translate to improved efficiency, cost savings for airlines, and enhanced passenger experience. While increased compliance reduced taxi delays, a slight decrease in runway throughput (3.2\%) was also observed. This suggests a potential trade-off between optimizing runway usage and minimizing delays. The study also reveals a spectrum of compliance among ATCOs, influenced by factors like experience and age. Qualitative feedback indicates high user satisfaction with I-MATE, suggesting its usefulness, reliability, and trustworthiness. This research underscores the value of AI-based decision support systems for air traffic control, particularly for addressing the cognitive challenge of balancing long-term outcomes with immediate actions, thereby paving the way for further advancements in airside traffic management.