In civil aviation, safety is paramount, and operations generate massive heterogeneous data from flight decks, towers, ramps, and maintenance, often at the network edge.
However, cloud-centric deployment of large Artificial Intelligence (AI) models often results in high task latency, lacks offline capabilities in communication-denied environments, and necessitates centralizing sensitive data, raising privacy and sovereignty risks. Edge AI brings perception, prediction, and decision-making closer to data producers through compression, collaborative inference, and split learning, thus reducing latency, bandwidth consumption, and exposure while allowing graceful operation during disconnections.
This paper provides a panoramic view and common understanding of edge intelligence tailored to civil aviation. It first articulates the operational motivations for edge AI, then reviews recent techniques for edge inference and edge learning.
Next, it introduces organizational computing paradigms and their respective configurations within civil aviation environments; finally, it describes emerging applications and future research trends in edge intelligence for civil aviation. It argues that a refined edge solution can complement cloud foundations to deliver low-latency, privacy-preserving, and resilient AI services across the civil aviation lifecycle.
Blogger's Review: This paper delves into the necessity and applications of edge intelligence in civil aviation, highlighting its potential to enhance safety and efficiency. As data volumes surge, edge computing is poised to become a crucial direction for future developments in aviation technology.