ABSTRACT: The next frontier of space exploration demands a new class of intelligent, autonomous systems capable of independent decision-making in unknown, dynamic environments. While Artificial Intelligence (AI) and Machine Learning (ML) offer transformative promise, their direct application in space is hindered by a unique set of severe challenges. Missions must contend with extreme constraints on power and computation, a scarcity of a priori training data for novel environments, and the uncompromising reliability and mission assurance standards of spaceflight.This talk will explore how NASA’s Jet Propulsion Laboratory (JPL) Robotics is confronting these challenges head-on. We will discuss the ongoing research to develop and infuse pragmatic, next-generation AI/ML techniques that are resource-aware, data-efficient, and verifiably robust. Finally, we will highlight how these trustworthy autonomy solutions are beginning to be deployed, bridging the gap between terrestrial AI and the real-world demands of robotic space exploration.

BIOSKETCH: Dr. Amir Rahmani is the supervisor of the multi-agent autonomy group at the NASA’s Jet Propulsion Laboratory (JPL), lecturer at California Institute of Technology, co-chair of the IEEE Space Robotics Technical Committee, and an associate fellow of AIAA. Additionally, Amir serves as the NASA SBIR subtopic manager of several portfolios on AI, autonomy, and multi-agent systems. With over two decades of research experience, Amir specializes in distributed space systems and swarm robotics. He is currently overseeing several projects focused on autonomy technologies for planetary and terrestrial robot teams.