Across six key domains — safety, operations, travel, logistics, regulation, planning & construction — a systematic map of "AI + Civil Aviation" scenarios: where AI applies, what it solves
Proactive risk sensing, precise early warning and rapid response across flight safety & maintenance, airport operations & security, ATC risk alerting & emergency response, and general & low-altitude aviation safety.
Multi-source data fusion, situational awareness & prediction, and dynamic strategy optimization for flight scheduling & dispatch, airport turnaround & resources, ATC command & capacity, and multi-party coordination.
Precise recommendations, trip planning, revenue optimization, frictionless border crossing and smart guidance — immersive intelligent service for passengers.
Computer vision, smart scheduling and embodied intelligence for digital logistics integration, efficient air cargo operations and reduced-crew support — lower costs, better quality.
Applying AI to safety oversight, market regulation and digital government to build a smarter, more precise regulatory system.
Smarter decision-making, precise control and dynamic optimization for airport siting, route networks and infrastructure lifecycle management.
Strengthening aviation AI foundations across data, platforms and models
Multi-party data integration with full-lifecycle dataset standards, building the corpora and knowledge bases for industry models.
Aviation computing infrastructure and industry data circulation facilities, built efficiently and operated securely.
Civil aviation LLM R&D, specialized vertical-scenario algorithms, evaluation mechanisms and a shared algorithm ecosystem.
From smart customer service to core operations, LLMs are reaching every link of civil aviation
21 TravelSky achievements selected as typical airport AI scenarios across safety, operations, services and general AI capabilities
Full-stack local deployment of DeepSeek R1-671B fusing 18 types of heterogeneous data — flights, logistics, IoT device logs
10 scenarios selected as national cases; DeepSeek + Qianrang enables real-time weather prediction
"Qianrang" — the first vertical LLM for civil aviation, providing specialized foundation capabilities
Full-stack local deployment of "DeepSeek R1-671B" with an airport-specific knowledge base, fusing 18 types of heterogeneous data — an AI breakthrough in airport safety.
"DeepSeek + Qianrang" for real-time weather processing and trend prediction with auto-generated alerts and operational suggestions — 3-hour weather forecasting.
DeepSeek integrated into the smart voice-inquiry system — multimodal AI understanding combined with the airport knowledge base: speech-to-text with efficient Q&A.
"DeepSeek + Qwen" mode tested in load-planning: 30+ airline manuals decomposed item by item to reduce hallucination for controllable, safety-critical use.
An ACMS DeepSeek knowledge base covering maintenance manuals, cases, experience, instructions and troubleshooting guides supports frontline engineers with intelligent fault diagnosis.
Local DeepSeek deployment with per-department custom training themes, dedicated accounts and workspaces — smarter, better-coordinated ATC.
Steady phased progress for high-quality "AI + Civil Aviation" development