Edge AI Transforms Civil Aviation Operations for Safety, Efficiency

Civil aviation operations, from flight decks to maintenance, generate vast amounts of data at the network edge, which traditionally relies on cloud-centric AI. However, this approach often leads to high latency, lacks offline capabilities in disconnected environments, and poses privacy and data sovereignty risks due to centralized sensitive data. This paper explores the concept of Edge AI, which deploys perception, prediction, and decision logic closer to where data is generated. Edge AI achieves this through techniques like data compression, collaborative inference, and split learning, thereby significantly reducing latency, bandwidth consumption, and data exposure. It also enables graceful operation during communication outages. The research provides a comprehensive overview of Edge AI tailored for civil aviation, detailing its operational motivations, reviewing current edge inference and learning techniques, and outlining organizational computing paradigms and emerging applications. The authors argue that refined edge solutions are crucial to complement cloud foundations, delivering low-latency, privacy-preserving, and resilient AI services across the entire civil aviation lifecycle.