Advancing intelligent water systems through new methods in model discovery 

I'm a postdoctoral researcher with Prof. Branko Kerkez at the University of Michigan in Ann Arbor. Our lab maintains a wireless water level sensor network (150+ devices) across the state that serves utilities, transportation authorities, and conservation organizations.

My research focuses on model discovery in low observability environments. In other words: how do you design a model or controller for a system with very few sensors? 

The primary impacts of my dissertation work were

• enabling a statewide water level prediction network at a tenth the cost of existing approaches

• enabling state space control of stormwater systems at scale

• a new system identification method which is natively continuous-time and maintains meaningful state naming.