DARPA’s plan for an airborne COVID detector
With COVID-19 primarily spreading through the air in enclosed spaces, the Defense Advanced Research Projects Agency wants a way to identify SARSCoV-2 signatures indoors and use that data to build a prototype sensor that can accurately detect the virus in the air quickly enough to stop its spread in office buildings or schools.
The small and variable characteristics of the virus combined with complex indoor environments make using a single detection and measurement technique extraordinarily difficult. Current COVID detection requires capturing a sample and sending it off to a lab for genetic analysis – a process that can take days. Optical environmental sensors, which can offer fast detection times, are not always able to discriminate between benign and pathogenic material.
“Current methods are not suitable for room-sized, indoor environmental monitoring,” DARPA said in a presoliciation. They “lack practical combinations of sensitivity, specificity (precision and recall), acceptable false positive rates, and speed and/or have substantial barriers to scaling due to cost or size, weight, and/or power requirements.”
The SenSARS program aims to overcome these existing challenges to environmental monitoring. DARPA suggested that recent developments in radiofrequency vibrometry, sensors, mass spectrometric techniques, immunosensing, electrochemical detection and machine-learning-powered signal analysis might help detect low concentrations of the virus.
DARPA is primarily interested in three use cases: detecting the virus in a 50-cubic-meter office, similar detection in a 300-cubic-meter conference room or classroom, and central monitoring of HVAC systems in buildings up to 10 stories. Solutions must have size, weight and power requirements suitable to the use case, and they must also inexpensive and easy to use and maintain.
Of secondary importance, DARPA said, is the ability to detect other pathogens within a month of discovery and the capacity to save samples for additional analysis.
The two-phase program is expected to produce three working prototype sensors in 18 months.
Proposals are due Dec. 1.
Connect with the GCN staff on Twitter @GCNtech.