Modeling quarantine during epidemics by mass-testing with drones

Leonid Sedov; Alexander Krasnochub; valentin polishchuk

Epidemic spread and drone technology are topics of broad research and public interest (the former heated by the current events, while the latter driven by the latest technological developments in the drone industry). We study the potential of using drones to deliver tests, enabling mass-testing for the infection. We show how the testing intensity, increased with the use of drones, may "flatten the curve" by quarantining test-positive population --for that, we extend the classical SIR epidemic spread model, introducing the "quarantined" compartment and solving (numerically) differential equations which govern the extended model. Importance of proactive COVID-19 screening, including symptomless people, has been acknowledged [BYWT+20, FLNI+20, HSXJ+20, LPCS+20, NKMS+20, WNB20]. However, mass-testing may be seriously impeded by population's fear of visiting testing facilities due to potentially high concentration of infection there; the fear is confirmed by health officials who advise against visiting hospitals "unless necessary" (examples of such directives from authorities during COVID-19 pandemic abound): "Stay home" is the overarching recommendation during pandemic virtually everywhere in the world. The good news is that COVID-19 test does not have to be necessarily conducted at a designated facility because (despite being somewhat unpleasant,) the test can be self-administered: a person may collect the material him/herself or with the help of a family member. Still, having people go somewhere to pick up and drop the tests would beat the purpose of the social distancing. A possible solution is to use drones to distribute tests to the population as well as to collect the tests back, bringing them to laboratories; the test results could then be communicated back to people electronically, so that those with positive tests put themselves into quarantine. Here we follow the infection dynamics by extending the SIR model [KM27] to include the compartment for quarantined population and show how the testing intensity, increased with use of drones, decreases the epidemic spread. Since pre-symptomatic infectiousness makes early surveillance and control crucial (symptom-based actions are not as effective), our results may guide the cooperation between health and transportation authorities towards optimal use of the available resources.