Ali Kashani is a postdoctoral research associate in the University of Virginia School of Data Science. He develops glucose control algorithms in collaboration with the UVA Center for Diabetes Technology to enhance the safety, reliability, and real-time performance of artificial pancreas devices.
Kashani holds a Ph.D. in Mechanical Engineering from the University of New Mexico. His research is focused on data-driven constrained control of nonlinear systems, combining physics-informed machine learning with formal safety guarantees from modern control theory. He developed model predictive controllers and probabilistic invariance techniques for safety-critical autonomous systems, including self-driving cars, autonomous drones, HVAC systems, and lane-keeping and soft-landing problems.
In addition, Kashani's work also introduced novel algorithms for learning barrier functions and neural Lyapunov functions, enabling formal stability verification and constraint enforcement in unmodeled, complex nonlinear environments. Kashani has contributed to improving extremum seeking control efficiency through dynamic optimization methods, designed scalable model predictive control strategies for large multi-agent systems, and developed adaptive controllers for Delta robots.