About Workshop
Aim
What Participants Will Learn
-
Frame AI/adaptive RL for hypersonic flight within a safety-first, non-weaponized context.
-
Build and document hazards, oversight roles, abort criteria, and geofencing for high-speed test articles.
-
Specify perception and state-awareness requirements without exposing sensitive control algorithms.
-
Draft a V&V plan, test envelopes, and safety monitors for AI-enabled high-speed systems.
-
Outline a safety case and operator SOPs aligned with governance and certification thinking.
Structure
📅 Day 1 – Safety, Ethics & High-Speed Flight Basics (Non-Weaponized)
- Where AI fits in safety-critical aerospace: hazard analysis, “human-on-the-loop,” and oversight.
- High-level aerothermodynamics concepts (no design details): why high speeds amplify uncertainty and sensing challenges.
- Assurance artifacts: requirements traceability, safety cases, and model/system cards.
- Hands-on: Build a policy-aware requirements matrix and hazard log for a benign high-speed test article (e.g., generic aero model) focusing on oversight, abort criteria, and geofencing.
📅 Day 2 – Robust Perception & State Awareness (Conceptual)
- Sensor integrity at high dynamic pressure: fault concepts, latency awareness, and graceful degradation (no algorithms).
- Observability at a glance: what it means to estimate states safely without revealing implementation.
- Validation & test planning: scenario coverage, limits, and transparency for operators and regulators.
- Hands-on: Draft a verification & validation (V&V) plan: define test envelopes, safety monitors, and operator intervention thresholds for a conceptual high-speed vehicle.
📅 Day 3 – Governance, Certification & Human Oversight
- Standards and certification thinking for AI components (conceptual): documentation, audits, incident reporting.
- Envelope thinking without controllers: defining stay-out zones, rate limiters, and conservative defaults.
- Human-machine interfaces: alerting, explainability for operators, and abort workflows.
- Hands-on: Assemble an assurance case outline (safety case) with roles, evidence to collect, and an operator SOP for safe testing and shutdowns.
Important Dates
Registration Ends
Workshop Dates
What You Will Gain
Outcomes
-
Apply a safety-first, non-weaponized framework for AI/RL in hypersonic flight.
-
Create core assurance artifacts: requirements matrix, hazard log, V&V plan, and safety case outline.
-
Define safe envelopes, stay-out zones, geofencing, and abort criteria for high-speed tests.
-
Design human-on-the-loop oversight with clear alerts, intervention thresholds, and shutdown workflows.
-
Navigate governance and certification thinking for AI in sensitive aerospace systems.
Who Should Attend
-
Engineers & researchers in aerospace, mechanical, electrical, controls, or related fields
-
AI/ML & RL practitioners interested in safety-critical aerospace applications
-
Flight test, safety, governance, and certification professionals
-
Senior students, faculty, and industry R&D teams working on hypersonic or high-speed systems