AirCascades: A visual analytics system for cascading taxiway interaction patterns

doi:10.1016/j.visinf.2026.100335

AirCascades: A visual analytics system for cascading taxiway interaction patterns

2026-06-11

0

OA

AI

DOI：10.1016/j.visinf.2026.100335

Original

Save

View PDF

Abstract

En 中文

Efficient management of aircraft taxiing operations depends on understanding not only where local conflicts occur, but also how they propagate through the taxiway network and evolve into cascading delays and operational risks. Existing research and operational tools still focus primarily on pairwise conflicts, raw trajectories, or coarse aggregates, making these propagation mechanisms difficult to inspect. To address this gap, we present AirCascades, a visual analytics system for post-hoc analysis of cascading taxiway interaction patterns. AirCascades uses a semantics-driven abstraction pipeline that transforms trajectories into encounter sessions, links them into cascade instances through shared-aircraft continuity, resource overlap, and topology-aware coupling cues, and clusters these instances into motifs that expose recurring propagation structures. Building on this hierarchy, the interface provides a macro-level view for system-wide screening, a meso-level pattern view with map-free glyphs for structural comparison, and a micro-level replay view for tracing individual cascades through speed profiles and inter-event links. A coupling-sensitivity analysis, a structural audit of the default analysis set, and qualitative expert walkthroughs on 24 h of operational data from a major international airport suggest that AirCascades helps analysts inspect propagation mechanisms and recurring structural bottlenecks that conventional surveillance replay does not readily expose.

Keywords:

Visual analytics

Spatiotemporal data

Airport operations

Interaction cascades

Pattern mining

AirCascades: A visual analytics system for cascading taxiway interaction patterns

Abstract

AI Summary

Journal

Organization