Since the onset of the pandemic, I’ve been working with Ranveer Chandra, Chief Scientist for Microsoft Azure Global, and his team to help conceptualize and convey whether and to what extent COVID19 related illnesses pose a risk to agricultural and food production. Today, we finally have a beta version of our dashboard ready for viewing.
The basic idea is to compare the geographic pattern of COVID cases to the geographic distribution of agricultural workers and agricultural production for particular crops and animals. If there are a large number of COVID cases in a location that grows the vast majority of a particular agricultural crop, then we can infer that production of that crop is at risk. The level of risk is inferred by creating an estimate of the percent of total state or national production of a crop that is potentially lost from COVID illnesses.
As it turns out, for the handful of agricultural commodities we’ve entered in our tool thusfar, the risk is very low (far less than 1% of total production estimated to be at risk). The reasons are straightforward: 1) production of most major agricultural commodities is distributed over a wide geography, and 2) the percent of the population with COVID in rural/agricultural areas remains low.
One of the main purposes of the tool is to help people visualize a portion of our food supply chain, to help people better understand where their food comes from, and to help illustrate that, at least at the moment, COVID pose little risk to the aggregate supply of food in the United States.
There are several ways I hope to expand this tool, should funding emerge and time allow. As we’ve seen, the risks from COVID appear to be less related to agricultural production and more related to food processing. Data on location and distribution of food processing facilities, and numbers of food processing workers employed in each location, is harder to come by but it’s not an impossible task. There are other items we’ll continue to work on to help make the tool more user friendly and functional.
A screenshot of the dashboard is below, but you’ll need to go to the Purdue page to actually use the tool.
Thanks to Ranveer Chandra, Hope Foley, Riyaz Pishori, Anirudh Badam, Jerry Neal, Stacey Wood, Peeyush Kumar, and Deepak Vasisht from Microsoft and Aaron Walz and Kami Goodwin at Purdue for their help getting this dashboard up and running.