We are living in the burgeoning age of machine learning and artificial intelligence, underlying real systems (e.g.robots and self-driving vehicles), and virtual systems (e.g. financial and inventory management). However, many of these autonomous systems have become so intricate and black-box that we face a complexity roadblock. For example, it can be difficult to formally explain why a classifier or a recommendation engine based on machine learning works. In addition, when the algorithms do work, how can we quantify their limitations and trust that they will perform effectively as intended. My research over the past has focused on using tools from controls, formal methods, optimization, and information theory to provide assurances for these autonomous systems, with application to robotics and aerospace systems. My research has significantly benefited from collaborations with industry partners, such as Raytheon, and researchers at Air Force Research Laboratory (AFRL) and NASA Jet Propulsion Laboratory (JPL). Below, I briefly describe four problem areas in autonomy that I study, namely, risk-sensitivity, safety, perception, and security.
