This thesis combines the practical development of an AI-assisted landing system with safety assurance inspired by ISO/PAS 8800. Using Microsoft Flight Simulator, the student will build a data collection pipeline, collect and curate flight data, develop or integrate an AI component, and implement closed-loop automatic landing. The system will be evaluated under varying operating conditions, with evidence linked to explicit safety requirements.

How can ISO/PAS 8800 guide the development and evaluation of an AI-assisted landing system? The thesis will investigate how to translate safety objectives into data requirements, implementation decisions, and measurable acceptance criteria. It will assess landing performance and failure cases within a defined operating domain and identify limitations in transferring automotive AI safety principles to avionics.

• Committed to producing research suitable for scientific publication.
• Good coding skills, including effective use of AI coding assistants.
• Able and willing to build simulator interfaces, collect and process data, and implement and debug landing control.
• Hardworking and willing to turn ambiguous problems into concrete research tasks.
• A background in formal methods, control, or aviation is a plus.