Drone services company Draganfly is evaluating a sensor package and artificial intelligence software that could be used to screen large area crowds for symptoms of dangerous contagions such as the coronavirus. The company already has received a “serious inquiry” from a potential, unidentified customer for this service package, according to CEO Cameron Chell.
Chell told AIN that the sensors would be effective at ranges up to 100 feet, flying drones indoors or outdoors at large venues including large sports stadiums, ports of entry, and refugee camps. Such a sensor package would include infrared, doppler radar, and high-resolution cameras to feed imagery to software that would be able to discern stress recognition, fevers, watery eyes, and muscle movement indicative of high blood pressure and present an initial analysis within 15 to 20 seconds of overflight.
To be optimally effective, the lag between overflight and data analysis would need to be cut to five seconds, Chell acknowledged, calling that “not a huge technical problem.” He said that, while facial recognition software could be added to the data package, the purpose of the flights would be to identify the pervasiveness of likely infection within a given population, as opposed to isolated specific individuals. Suitable hardware platforms include Draganfly’s Commander quadrotor or popular commercial, off-the-shelf models including the DJI M600. Indoor flights could be accomplished with ducted fan aircraft to allay safety concerns. “These are a bit bigger than your typical hobbyist drone, but they can still be small enough to fit inside a backpack,” Chell said, stressing that the technology is relatively easy to deploy.
The advantage of using drones for a public health application is the much larger area they can cover and the increased number of individuals who can be sampled in a small block of time, as opposed to using stationary sensors. Drones can also be more cost-effective.
“In many places, you don't have the luxury of putting in a hard [sensor] installation, putting up walls, or erecting checkpoints. Those things are incredibly invasive and stressful to a population. Putting a drone in the air allows you to monitor crowds and know exactly what, and sometimes where, a population has been exposed,” he said. In the aggregate, Chell said that a single drone overflying a crowd can collect an “enormous” amount of data over a very short period of time. “In 10 minutes you can identify so many datasets,” he said. Chell said a typical public health sensor package could be fitted as an add-on to existing public safety drones and integrated into current public safety unmanned flight programs.
Using drones for population health sampling also can provide the critical data medical decision-makers need when deciding where and how to deploy resources in an emergency such as the coronavirus outbreak, Chell said. The system can use Draganfly’s software in the cloud and also could interface with the end-user’s software. The nuts and bolts of that arrangement are comparatively less important than providing the data as expeditiously as possible, Chell said. When it comes to the most accurate data on contagions like the coronavirus, health care institutions and policymakers “just want the result.”