AI-supported design and autonomy in aviation
Generative design, surrogate models and materials discovery, certification evidence, the autonomy roadmap, detect and avoid, assuring learning systems, workforce and liability, and how to judge a claim. 2026 edition. Part 4 of the AI in aviation series.
What you will be able to do
- ✓ Distinguish AI used as a design tool from AI inside the flying product, separate automation from autonomy, and label claims as in service, certified, demonstrated, research or proposal.
- ✓ Explain how generative design, surrogate models and computational materials discovery are used, and what evidence (validated analysis, tests, tool qualification, model credibility) turns their output into certification evidence.
- ✓ Describe the verified status of autonomy in 2026, from airliner pilot aids and reduced-crew research to uncrewed cargo and certified pilotless passenger flights, and build a gated autonomy roadmap.
- ✓ Explain detect and avoid, the difference between learned and learning AI, and run-time assurance, using the official findings on the Docklands drone swarm accident (ATSB AO-2023-033).
- ✓ Assess workforce, supervision and liability effects of AI and autonomy, including Europe's product liability rules from December 2026.
- ✓ Evaluate a vendor's AI or autonomy claim with five questions and events-per-year arithmetic, and present maturity and a staged business case to a board.
Course content
Session 1 · The frontier: AI in design and in flight
1.1
- Two frontiers, five status labels · free preview
- Generative design and evolved hardware
- New materials and surrogate models
- Worked example: Tolimán Aero's frontier map
Session 2 · AI in the design office: evidence that certifies
1.2
- Analysis, test and the proof of structure
- When a tool's output becomes evidence
- Is the model credible here?
- Worked example: cross-checking a vendor's surrogate model
Session 3 · The autonomy roadmap: from cockpit aids to pilotless flight
1.3
- Pilotless passengers, and the autonomy ladder
- The airliner path: taxi, take-off, landing and crew
- Uncrewed cargo and air taxis
- Worked example: a gated autonomy roadmap
Session 4 · Detect and avoid, and assuring systems at the frontier
1.4
- Replacing the eyes: detect and avoid
- Learned and learning systems
- Run-time assurance: a simple guard for a clever function
- From the investigation file: the drone swarm that lost the wind, Docklands 2023
Session 5 · People, work and liability
1.5
- Jobs: what changes, what stays
- Liability: who answers when software decides
- Supervising automated fleets
- Worked example: staffing a remote operations centre
Session 6 · Judging AI and autonomy claims, and what comes next
1.6
- Five questions for any claim
- Reading maturity: readiness levels and certification progress
- The business case and where to invest
- Worked example: judging a detect-and-avoid claim, and the road ahead
Who this course is for
Course content © Aerospace Garage Limited. All rights reserved.