This research explores how large language models can support complex problem formulation in mission engineering contexts, with a focus on NASA space mission design challenges.

• Study documented ChatGPT's performance variability across multiple prompting attempts
• Evaluated AI's ability to translate ill-defined problems into well-structured engineering requirements
• Identified key strengths and limitations of current LLMs for engineering problem formulation tasks
• Provides insights for engineering practitioners on effective LLM integration

This work offers valuable guidance for engineering organizations seeking to leverage AI tools while understanding their current capabilities and limitations in complex systems engineering applications.

An Empirical Exploration of ChatGPT's Ability to Support Problem Formulation Tasks for Mission Engineering and a Documentation of its Performance Variability