The algorithm, revealed Monday, works by combining user-input data on sleep schedules and maximum caffeine allowance with psychomotor vigilance task (PVT) tests. As it gathers data over time and gets to know the user, the algorithm can predict exactly when and how much caffeine they should drink in order to be more productive. It can even calculate how to get optimal results while reducing overall caffeine intake.
“We found that by using our algorithm, which determines when and how much caffeine a subject should consume, we can improve alertness by up to 64 percent, while consuming the same total amount of caffeine,” Jaques Reifman, a biotechnology software researcher with the U.S. Department of Defense, told Science Daily. “Alternatively, a subject can reduce caffeine consumption by up to 65 percent and still achieve equivalent improvements in alertness.”
Currently the algorithm is only being used on soldiers in training, but the Army is planning to license it as a smartphone app for the general public in the future.