Automation in the form of feedback control can have an important impact on biomedical systems. For example, during vascular surgery an anesthesiologist is responsible for monitoring blood pressure and cardiac output and adjusting the infusion rates of several drugs. In essence, the physician is serving as a feedback controller. I will present an automated strategy (multiple model predictive control) that we have developed and show experimental results (from canine studies) to demonstrate that this approach results in less variability in blood pressure and cardiac output than the current practice of manual regulation.

A recent focus of our research involves continuous glucose monitoring and control for diabetes. Individuals with Type 1 diabetes must closely monitor their blood glucose levels by pricking their fingers several times each day to obtain blood glucose values from glucose test meters. The availability of a continuous glucose sensor has the potential to dramatically improve blood glucose regulation and is necessary for a closed-loop artificial pancreas. I present an optimal estimation theory-based approach to infer blood glucose and its rate of change from noisy subcutaneous sensor signals. A hypoglycemia alarm can then be signaled to inform an individual of the need to consume sugar to avoid a potentially dangerous low blood glucose condition. Further, I present results on the detection and estimation of an “unannounced” meal, which can be used to provide feed forward action in a closed-loop artificial pancreas.

Finally, I discuss applications of systems and control that are related to a pharmaceutical industry initiative on process analytical technologies (PAT). Dynamic simulation is used to understand the impact of design on the operability characteristics of a pharmaceutical pilot plant. In addition, run-to-run techniques, commonly used in semiconductor device manufacturing, are extended and applied to a chromatographic separations process.