Explainable Reinforcement Learning for assisting Air Traffic Controllers

2026-07-24

AI Summary

This study addresses the challenge of limited explainability in AI decision-making within high-stakes domains such as air traffic control, where opacity undermines human trust and collaboration. For the first time, explainable artificial intelligence (XAI) techniques are integrated into a reinforcement learning system tailored for air traffic management. The agent is trained in a simplified air traffic simulation to plan routes that avoid no-fly zones, while saliency maps are employed to visualize the critical input features driving its decisions. This approach not only achieves effective route planning but also provides human operators with intuitive, interpretable justifications for AI-generated recommendations, thereby significantly enhancing trust and collaborative efficiency.

Abstract

To effectively integrate AI into high-stakes, critical environments such as healthcare, autonomous driving, and aviation--and to advance toward higher levels of automation and seamless human-AI collaboration--building trust in AI-driven solutions is essential. Trust, in turn, is closely linked to the explainability of AI systems. The rapid advancements in AI across various domains have underscored the challenges of establishing trust, raising increasing interest in AI explainability even more when applied to deep learning. In this context, the present work aims to explore the application of explainability techniques to Reinforcement Learning (RL) algorithms, specifically within the safety-critical domain of Air Traffic Control (ATC). Using a simplified ATC environment as an initial testbed, an intelligent agent is trained with a reinforcement learning algorithm to make decisions on alternative flight routes that avoid no-fly zones. As a preliminary explainability approach, a saliency map is employed, providing insights into the input features that most significantly influence the agent's decision-making process.

Problem

• Explainable AI
• Reinforcement Learning
• Air Traffic Control
• Trust
• Saliency Map

Innovation

• Explainable Reinforcement Learning
• Saliency Map
• Air Traffic Control
• Human-AI Collaboration
• Trustworthy AI