We outline a potential approach to the development of future autonomic NASA Space Exploration Missions. The approach is quite different from other efforts to date, and is inspired by the Human Immunological System and exploits agent-based technologies to support autonomic features such as: self-managing, self-configuring, self-healing, self-optimizing, self-protecting, and finally two novel features of future NASA swarm-based missions: self-predicting and self-learning, which may help in extending levels of autonomy in future missions. The approach offers several detection mechanisms for recognizing Self and non-Self patterns. This sense of "Self " of the Immunological System is related to the ability to learn to recognize new kinds of threats. The hierarchical architecture based on agents offers a high flexibility for repairing, maintaining, configuring, and optimizing functions of the spacecraft in real time. © 2007 IEEE.