Where is collected field data stored?

[TylerT]

Hello,

I want to create a Field Map/App and want to understand the data patterns better.  Say, I publish a web map with ArcGIS Pro.  Where will the collected field data be stored?

Thank you,

Tyler

ArcGIS Pro 3.1.0

[clt_cabq]

That really depends a lot on how you store and publish your data. However, the simplest answer and the one that I suspect fits your use case is that you are publishing a data set to the web as a hosted web service. In that instance you may have a geodatabase stored on your network or locally that serves as a sort of schema or template for what gets published to the cloud. Once published, the published data and your local database are two completely separate instances of your dataset and do not interact. The data collected in the field will go to the database stored (hosted) in the ArcGIS cloud. You can bring that service into ArcGIS Pro directly, edit it, use it to update the local stored data, visualize it for mapping, etc.  An alternative is that you are working with an enterprise geodatabase that is published as a web service through your own local ArcGIS Enterprise system - these data 'live' on your organizations servers (on premises or cloud if that is where your infrastructure is) and editing in the field directly updates that source. There are advantages and disadvantages to both approaches.

[clt_cabq]

Also, I should have added that in both cases, that if your field crews are working in a 'disconnected' environment (i.e. not connected to the internet), the data each individual collects is stored local to the device they are using (phone, tablet, other mobile device) until they make an internet connection when that data will be uploaded either automatically through a manual prompt to the ultimate data storage. This is important to understand for times when data seem to go missing, people have troubles with syncing their devices (which they will at times), or other circumstances that prevent the sync from taking place.

[TylerT]

@clt_cabq;

Thank you.  I like the idea of a single authoritative source (not two completely separate instances).  It sounds risky to pass data back and forth from ArcGIS Pro client to the hosted server.  And if I understand correctly, first source is from ArcGIS Pro, then flow reverses there forward. 

Also, what if your schema changes after that first pitch?

I am working with Enterprise, but IT has yet to stand up SQL server for the EGDB, so I'm preparing for eventualities.  

Thanks again,

Tyler

[clt_cabq]

If you aren't working with Enterprise, I think the best practices would be to build out your schema as completely as possible, publish it as a hosted service in AGOL, and make sure you work with that as your authoritative data, and backing it up with some sort of regularity. If you change your schema part way along, you'll have to figure out the best way to make that change operationally, but this is going to be true if you are publishing a web service from Enterprise or if you work with hosted feature classes in AGOL. One thing to understand about hosted services or feature classes is that they are not likely to take advantage of some of the more sophisticated features of a file or enterprise geodatabase like attribute rules, contingent values and the like. Managing and maintaining domains can also be challenging with hosted feature classes, so if those things are going to change frequently then you will need to maybe skip those in your geodatabase and edit those in the hosted feature class. There may also be some different things to deal with if you are using Field Maps versus Survey123 for data collection. The biggest thing you need to do is really plan out what your data life cycle looks like and determine the best ways to handle many of these issues for the workflow you are building.