Research/ PLANETARY ROBOTICS

Field Test of a Mission Science Objective-Based Artificial Intelligence for Spaceflight

Bethany Theiling, Shannon Bull, Connor Williams, Leyton McKinney, Connor Firth, Lily A. Clough, Alan Gibson, James Marshall, Steven Hard, Michael Monaghan, Jonathan B. Hammer, Melissa De La Cruz, Caroline Zouloumian, Brett Allen McKinney, Evana Gizzi

IEEE CAI 2026/May 8, 2026

Space × AIPLANETARY ROBOTICS

Technique

Autonomy, Task Planning & OperationsOpenAlex topic: AI-based Problem Solving and Planning

Indexed as

OpenAlexAI-based Problem Solving and Planning · Artificial Intelligence Applications · Planetary Science and Exploration

DOI Semantic Scholar OpenAlex

Onboard AI-driven decision-making will be necessary for future spaceflight missions that explore increasingly challenging environments with limited communication to Earth. We developed and tested a science mission Objective-Based AI (OBAI) using uncrewed aerial systems (UAS) in a Mars and Enceladus planetary field analog site. OBAI is integrated into an end-to-end AI framework to ingest and analyze instrument data, make decisions based on onboard analyses, request action by secondary assets, and plan and execute a new mission: the Adaptive Sensing Technology for Responsive Autonomy (ASTRA) AI capability. We demonstrate mission-level AI decision-making in power-limited systems, enabling OBAI and ASTRA to be integrated into future spaceflight missions.