The Department of Defense’s (DoD) Defense Innovation Unit (DIU) is seeing good results on a program that uses AI combined with commercial off-the-shelf wearable-devices to identify if the wearer has contracted COVID-19 up to two days before symptoms show.
DIU Director Michael Brown touted the program, known as the Rapid Assessment of Threat Exposure (RATE), at the National Defense Industrial Association’s National Security AI Convention on March 23.
“We’re proving with a sample of about 8,500 folks that we can detect COVID up to 48 hours before you ever feel the first symptom,” Brown said at the event’s afternoon keynote. “[It’s a] critical capability if we think about a ship, with its crew going out to sea, it’s important to understand very early.”
The RATE program was initially started back in 2018, to help detect the SARS virus – another coronavirus – and pneumonia but pivoted last spring when the COVID-19 hit shores. The program received $7 million in funding through the CARES Act, according to a September 2020 DoD release, and has progressed to the point where Brown said DIU is almost ready to transfer it to the Defense Threat Reduction Agency.
Brown said the program – which uses wearables and an AI/ML algorithm to keep track of wearer’s temperature, pulse oximeter, and cardiac measures – is currently in use on the USS Portland, experimenting with its accuracy at sea.
Brown himself was wearing the testing devices at the event and said the project’s results have been “great” and DIU will transfer it in what is likely record time, or close to it.
Another DIU COVID-19 project, project Clean Sweep, is being pioneered with the Navy, Brown said. The project would utilize robotics and ultraviolet light, as well as disinfectants to be able to deep clean an enclosed space without potentially exposing humans to the virus. Brown said he would look at a report on Project Clean Sweep later that day.