Undershoot/Overshoot Geometric Error Fixes

[KokilaM]

Hi Team , 

We are migrating data from G Connect to Esri . Since the data have so many Undershoots /Overshoots , So many disconnected features in Utility Network . Do we have automated way to bulk fix these issue ?

@RobertKrisher @MikeMillerGIS @JoaquinMadrid1 @UserDistributor 

Thanks,

Kokila 

[RobertKrisher]

I'm not aware of any single tool that will automatically fix all these issues in a way that creates a usable topology. FME / Data Interop have many powerful tools for finding these situations and performing geometry manipulation to resolve them, there are also core Python and GeoProcessing tools you can use to identify them.

If you have a lot of these issues, or other, issues to fix you should consider reaching out to a distributor or business partner with data migration experience. Many of them have already developed models to identify and resolve these issues.

[gis_KIWI4]

@KokilaM - What @RobertKrisher suggested. 

If you use FME be careful about the tolerances when using transformers like snapper, it can cause unwanted "joins". This risk will always exist when getting a tool to make some decisions around editing the geometry. You can make it robust and lower this risk by adding rules into the the decision making. 

More than likely you will need a few iterations trying to fix different variations of the of issues. 
I would take the approach of finding all distinct types of overshoots and undershoots problems (certain assetgroups and asset types, presence of a device/junction, need to create a vertex?) and work out a few methods for fixing the majority. Hopefully that will bring it down to a manageable level that can be fixed manually.

[KokilaM]

@gis_KIWI4 you are absolutely Correct.  Its creating un relevant joins where the under shoot/Overshoot doesn't exists in FME with the Snapper tool

[gis_KIWI4]

@KokilaM - That's expected. In the above example there are not undershoots/overshoots, best to filter these off and then send the rest to the Snapper. 
Within Snapper you have a couple of options to snap to end points OR to snap to vertices. You might have both kind of scenarios in your dataset. IF you know the lines should only snap to a certain type of object/device then only send those and filter the rest. 

But the short answer here is that it will all depend on the data and the solution devised will be custom to the data issues you have.

[RobertKrisher]

To add onto what @gis_KIWI4 said, you don't want to just pass everything into an anchored snapper. That would be a mess. What you want to do is use transformers to identify the problem, and apply a targeted fix precisely.

Mature data migration vendors have already worked out the logic for doing this reliably and safely. When done improperly, this can lead to topology errors (self-intersecting lines, duplicate vertices, etc) and disconnected features (stacked points, missing vertices, etc). But if you don't have too many of these you and/or you have enough free time and grit then it is a tractable issue.

[KokilaM]

Thank you @RobertKrisher @gis_KIWI4  . I am working on the solution to find out the areas where undershoot/Overshoot exist using an FME , These would be around 0.4 Million fixes . So we can't trace all the issues .