One of the major reasons why the artificial pancreas remains a possibility rather than a reality is because it’s hard to create a closed-loop system that can account for all the variabilities of day-to-day life. Developers have been trying to figure out algorithms that factor in the body’s reaction to hard-to-quantify stimuli like carbohydrate counts of meals and the strenuousness of exercise. And if a pump system can’t factor in such things, it falls short of being a closed-loop artificial pancreas.

Researchers are launching a major new study to try to solve least one missing part of this equation. The Illinois Institute of Technology, the University of Illinois, the University of Illinois Chicago, and York University (Canada) announced they have received a $2.48 million grant from the National Institutes of Health to develop artificial pancreas systems that can handle the variability of exercise without additional input from people with diabetes. The goal of the 5-year study is to develop an artificial pancreas that has a built-in early warning system for hypoglycemia brought on by physical exertion. The team of researchers have been working on AP systems that can use algorithms and physiological data to “learn” to predict upward or downward trends of BG levels in response to stimuli.