Michael G. Safonov

University of Southern California, Southern California University for Professional Studies

Papers

Total Citations

H-Index

About

Michael G. Safonov is a leading figure in control theory, best known for pioneering the field of **unfalsified control**—a data-driven approach that challenges traditional model-based adaptive control. His key research areas include robust control, adaptive systems, and nonlinear dynamics, with a focus on developing controllers that can learn and adapt without requiring an accurate mathematical model of the system. Safonov’s major contribution is the **unfalsified control framework**, where candidate controllers are eliminated in real-time when their performance is "falsified" by incoming data, rather than being tuned based on a potentially flawed model. This paradigm shift enables safer, more reliable adaptive control for complex systems. His most cited work, "Unfalsified direct adaptive control of a two‐link robot arm" (2001, 28 citations), demonstrates this theory on a nonlinear robot manipulator, showcasing its practical viability. Safonov’s ideas have had a lasting impact on aerospace and robotics, where uncertain operating conditions and unanticipated failures demand robust, learn-on-the-fly solutions. His work continues to inspire researchers in data-driven control and adaptive systems.