MAIFormer: Multi-Agent Inverted Transformer for Flight Trajectory Prediction

2025-09-25

AI Summary

For joint trajectory prediction of multiple aircraft, it is essential to simultaneously model individual temporal dynamics and complex inter-aircraft interactions while ensuring prediction interpretability. This paper proposes MAIFormer, a novel multi-agent Transformer architecture featuring a dual-attention mechanism: (i) masked multivariate attention captures intra-aircraft evolution across multidimensional states; and (ii) agent attention explicitly models airspace coordination and conflict-avoidance relationships among flights, enabling decoupled representation of behavior and interaction. Evaluated on real-world ADS-B data, MAIFormer achieves significant improvements over state-of-the-art methods in RMSE and MAE. Moreover, its attention weights provide intuitive, interpretable insights into influential aircraft and critical temporal steps, offering transparent, traceable decision support for air traffic controllers—thus balancing high prediction accuracy with operational utility in air traffic management.

Technology Category

• Multiagent Systems: Multiagent Planning
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Agentic search

Abstract

Flight trajectory prediction for multiple aircraft is essential and provides critical insights into how aircraft navigate within current air traffic flows. However, predicting multi-agent flight trajectories is inherently challenging. One of the major difficulties is modeling both the individual aircraft behaviors over time and the complex interactions between flights. Generating explainable prediction outcomes is also a challenge. Therefore, we propose a Multi-Agent Inverted Transformer, MAIFormer, as a novel neural architecture that predicts multi-agent flight trajectories. The proposed framework features two key attention modules: (i) masked multivariate attention, which captures spatio-temporal patterns of individual aircraft, and (ii) agent attention, which models the social patterns among multiple agents in complex air traffic scenes. We evaluated MAIFormer using a real-world automatic dependent surveillance-broadcast flight trajectory dataset from the terminal airspace of Incheon International Airport in South Korea. The experimental results show that MAIFormer achieves the best performance across multiple metrics and outperforms other methods. In addition, MAIFormer produces prediction outcomes that are interpretable from a human perspective, which improves both the transparency of the model and its practical utility in air traffic control.

Problem

• Predicting multi-agent flight trajectories with complex interactions
• Modeling individual aircraft behaviors and social patterns simultaneously
• Generating explainable prediction outcomes for air traffic control

Innovation

• Uses masked multivariate attention for spatio-temporal patterns
• Employs agent attention to model social interactions
• Produces interpretable predictions for air traffic control