Multi-modal Air Traffic Prediction
Background: Air traffic dynamics are influenced by complex factors beyond current aircraft states, including environmental conditions and human-machine interactions. Integrating these heterogeneous data sources into predictive models is a recognized challenge in the field.
Question / Future Work: There is an open need to incorporate external, multi-modal data—specifically meteorological conditions and voice communication transcripts—into state-to-flow predictive architectures to capture drivers of disruptions that purely kinematic models miss.
Why It Matters: Incorporating external variables is critical for robustness in volatile operational environments where state snapshots are insufficient.
Evidence: Future work may further incorporate external factors, including weather conditions and richer spoken instructions, to further enhance predictive performance under rapidly evolving traffic situations.
Metadata & Links
- created_at
- 2026-05-14T06:10:14Z