Coronavirus Dashboard Data Automation Methods

[Min-HungChou]

Johns Hopkins University's well-known Coronavirus dashboard had automation support from the Esri Living Atlas team, according to its description. I am interested in knowing what did the team do in regard to that? Since it is not a hosted feature layer and they have switched storing the data from Google Sheets to GitHub, how can a feature layer gets updated if the data is not uploading from a cloud service (Google Drive, OneDrive, Dropbox)?

And what are the differences between their own ncov_cases feature layer and the Living Atlas Coronavirus COVID-19 Cases layer?

I am fascinated with how Esri was able to help aggregating multiple data sources into something that the University can use. I would like to know more about their workflow.

Thank you

[DanPisut · Accepted solution]

Hello Min-Hung, 

The COVID-19 layer is a feature service that uses the same Aggregated Live Feeds methodology as many of the near real-time environmental layers found in Living Atlas. In short, we are downloading the latest data (from JHU), processing it locally, and publishing it as a hosted feature service. Every 20 minutes, the process is restarted and the service is overwritten with the latest information. 

The two layers that you linked are actually the same. JHU added additional information to the ArcGIS Online content item for inclusion in Living Atlas, but if you look at the REST, they are the same item ID. 

If you're interested in developing your own automated layers, there are two Learn ArcGIS lessons on implementing Live Feeds:
Schedule Automated Near Real-Time Data Updates | Learn ArcGIS 

Update Real-Time Data with Python | Learn ArcGIS 

Cheers,

Dan