Increase Survey123 accuracy

1709
9
08-29-2018 11:57 AM
MarkLache1
New Contributor III

I have a Survey123 web form with a geopoint. Right now I'm passing in coordinates via the URL scheme with over 10 decimal places. By the time the Survey123 web form gets it the coordinates appear to be truncated down to 5 decimal places. Any way to increase the decimal places?

Thanks

0 Kudos
9 Replies
JamesTedrick
Esri Esteemed Contributor

Hi Mark,

Currently, the level of precision for the webform is set to 5 decimal places; this translates to approximately an area of 1 square meter (~1.113m x ~0.787m at 45 degrees of latitude). Each decimal point increase beyond that will shift the decimal point to the left by one; at 10 decimal points, you're below .01 inches for each of those measures.  

Can you provide some details for the workflow you have that that uses the level of precision you are sending?

0 Kudos
MarkLache1
New Contributor III

Users are clicking on interior features of buildings and the shift from reduced precision will cause confusion and incorrect information to repair staff. Where they click should be where the point ends up. Wont need all decimal places, but will I need a few more decimal places. I find it weird how you cant control the level of precision in Connect.

0 Kudos
JamesTedrick
Esri Esteemed Contributor

Thanks Mark,

We can look into increasing the level of precision supported by the web form; based on your requirements, it sounds like 7 decimal places (which would set the rounding error to ~1.113cm x ~0.787cm at 45 degrees of latitude) would be adequate, yes?  For reference, the ArcGIS Online base maps (and the Campus Basemap template) support a maximum zoom level that is equivalent to 1 pixel corresponding to a 1.8cm (.71 inch) square.

0 Kudos
MarkLache1
New Contributor III

Sure lets test that. How do I accomplish this?

Thanks

0 Kudos
JamesTedrick
Esri Esteemed Contributor

Hi Mark,

I've logged an issue for our development team - precision is a fixed setting in the code.  It'll take a bit of time to make the code change and post it our beta testing environment.

As an alternative,  in the Early Adopter Community, we've just posted the first on-premises installation guide - it would be possible to modify a version of that locally for testing, though it's a fairly large technical solution to test a small setting change.

0 Kudos
MarkLache1
New Contributor III

Can you suggest any workarounds for the moment? My users will get confused if they see their point moving after they have submitted the survey.

0 Kudos
AndrewSouthern
Occasional Contributor

Was there any movement on this issue?  I don't need the level of precision as the initial post but would like the ability to be able to adjust the XY precision for locations passed into a survey from an external source as well.

Thanks.

0 Kudos
JamesTedrick
Esri Esteemed Contributor

Hi Andrew,

This was put on hold temporarily as we evaluated the level of precision required by different customers and received by different devices (for both the field app and webform).

0 Kudos
KyleWikstrom
Occasional Contributor II

Hi @JamesM ,

I was very glad to find this content on GeoNet because we are experiencing this challenge at a couple local governments that are using web map Arcade expressions to pass feature point geometry attributes as URL parameters into Survey123 forms. In some cases, staff users are submitting reports using Survey123 for property parcels, and the rounding of the Arcade calculated coordinates from 10-12 digits down to 5 digits in Survey123 is producing an offset of significant consequence when the Survey123 point feature gets misplaced on top of a different property parcel. The workaround is for staff to verify the location of the point in the map before saving the survey and to manually update it. However, if multiple surveys are submitted for a single property where the staff user is manually updating the point location each time, the points are getting scattered across the property parcel over time, and is impacting the perception of data integrity within the department.

If you're able to provide some information on where this idea/issue has landed, if it will be updated in an upcoming version, and if you or others have other workarounds to improving the prevision of the mapped point when using URL center paramenters.

 

Thanks, James!

0 Kudos