Maani Ghaffari

Maani Ghaffari is an Assistant Professor in the Department of Naval Architecture and Marine Engineering and Robotics at the University of Michigan. His work focuses on robotics, autonomous systems, and the integration of applied mathematics with machine learning. He leads the Computational Autonomy and Robotics Laboratory, advancing research in sensor fusion, state estimation, and geometric control. Ghaffari teaches courses such as NA/EECS 568: Mobile Robotics and ROB 101: Computational Linear Algebra.

His research interests include robotic perception, planning under uncertainty, and invariant filtering techniques for navigation systems. He has developed frameworks like GPS-DRIFT for marine robotics and contributed to SLAM algorithms for surgical and underwater applications. Ghaffari's work emphasizes robustness in dynamic environments, leveraging Lie algebraic methods and equivariant neural networks.

Notable contributions include correspondence-free point cloud registration, Bayesian semantic mapping, and energy-based terrain modeling for legged robots. His research bridges theory and practice, with applications in surgery navigation, autonomous vehicles, and industrial anomaly detection.