January 30, 2025, Sydney Dolan, Technical University Munich

Postdoctoral Researcher, Technical University Munich
2:15 p.m. 
310 Kelly Hall

"Visionary Satellites: How Perception and Planning Drive Autonomy in Orbit"

Abstract: Space profoundly impacts daily life, shaping our understanding of climate, weather, navigation, and communication. With ever-growing satellite networks and data surpassing human processing capacity, advancements in multi-agent robotics and AI promise to transform space exploration and enable breakthroughs in areas like climate monitoring and in-space assembly. However, various obstacles to effective multi-satellite systems arise, including limited available onboard resources, unclear communication protocols, and highly illuminated and occluded imagery. In this talk, I will present my research aimed at advancing satellite autonomy, with the goal of creating AI-enabled distributed space systems that possess a deep understanding of their surroundings and can independently make decisions. The talk will focus on two key areas (1) Multi-Agent Planning, which focuses on characterizing what information is most useful for satellite coordination and (2) Neuromorphic Perception, which uses event cameras to track changes in motion in highly illuminated and dynamic scenes from space. Through these advancements, we strive to move towards a new era of satellite autonomy, empowering satellites to perceive, plan, and respond to the world around them.

Bio: Dr Sydney Dolan is currently a postdoctoral researcher at Technical University Munich, where she develops neuromorphic computer vision algorithms for space situational awareness. Before joining TUM, Sydney earned their PhD in Aeronautics and Astronautics from Massachusetts Institute of Technology. Her research interests encompass multi-agent reinforcement learning, graph-theoretic multi-agent coordination, and spacecraft perception and pose estimation.