Researchers at the Beckman Institute for Advanced Science and Technology will begin two new projects — on human performance optimization in U.S. war fighters — sponsored by the Defense Advanced Research Projects Agency.
One program will establish and validate third wave artificial intelligence methods for individualized performance optimization within U.S. war fighters and another is designed to improve how the war fighter identifies, measures, and tracks personalized biomarkers.
Teaching AI to Leverage Overlooked Residuals
As part of the AI Next Campaign, a team based at the Center for Brain Plasticity at Illinois was awarded $1 million for a project within DARPA’s “Teaching AI to Leverage Overlooked Residuals” program.
The program will explore whether third wave AI methods can be applied to develop personalized interventions that are tailored to the needs of individual war fighters. Research investigating how to optimize human performance has long appreciated that each person’s response to intervention is not the same. The goals, abilities, and needs of each individual may differ. Recent work has therefore focused on developing new methods to optimize human performance by taking individual differences into account.
The Illinois team is led by Aron Barbey, director of the Center for Brain Plasticity and professor of psychology, along with co-investigators Sanmi Koyejo, an assistant professor of computer science at Illinois; Adam Strang, a human performance research scientist at the Air Force Research Laboratory at Wright Patterson Air Force Base; and Elizabeth Whitaker, a principal research engineer at the Georgia Tech Research Institute.
“DARPA has pioneered the development of third wave AI methods that aim to replace highly specialized modeling approaches with new methods for general purpose problem-solving,” Barbey said. “Our interdisciplinary team will use these methods to establish a new modeling framework that can design personalized intervention protocols to optimize performance – taking each war fighter’s cognitive and biological characteristics into account to predict their performance on a broad range of training protocols within the United States military.”
Measuring Biological Aptitude
In a second DARPA sponsored project, Barbey and colleagues are collaborating with Crystal Jiang, a professor of applied genomics and her team at Lawrence Livermore National Laboratory, whose proposal of approximately $9 million was selected in the Measuring Biological Aptitude program. Barbey and Steven Culpepper, an associate professor of statistics at Illinois, will lead the cognitive and biological phenotyping test and evaluation team at the Center for Brain Plasticity.
“Our goal is to improve how the individual war fighter identifies, measures, and tracks personalized biomarkers and therefore to help them prepare more effectively for specialized roles in their military career.” Barbey said. “We will work closely with DARPA and our colleagues at Lawrence Livermore National Laboratory to evaluate biosensing protocols that are designed to precisely measure and monitor biological aptitude and will also examine whether this information can be applied to improve the accuracy of personnel selection within the military.”
According to a DARPA announcement on the MBA program, “If successful, the MBA program will address a loss in force readiness by enabling individual war fighters to prepare more effectively for specialized roles at the beginning of their careers. This will lead to higher quantities of candidates who are better prepared for military specialties.”
“Additionally, MBA technology will allow selection boards to achieve greater precision by eliminating inherent subjective biases from the candidate identification process, and thus recognizing persons who otherwise might not have been identified with current techniques.”