Invalid Geometry Checks on routes

[PrestonStafford2]

We did a data review check on invalid Geometry on our routes. It brings back that we have 41 "Not Simple" issues. My question is is what is "Not Simple" mean and is this a critical issue.?

[RyanKoschatzky]

For item 10 fixing, Nathan has provided the following workflow to us.

In order to fix these routes (and events), I would follow the workflow below: 

1. Turn off Roads and Highways.
2. Remove one of the vertices that are within tolerance on each centerline.
3. Turn on Roads and Highways.
4. Run Generate Routes on the routes that utilize the centerlines you editing in step 2.
5. Run Generate Events on the events that utilize the routes you updated in step 4.  It’s a GP tool, so you can put selection sets of definition queries to significantly speed up this process and only run on the impacted events.

I am working through testing this for us. If you just remove the vertex on centerline, as I had suggest for our users, apply updates will become active, after running you have the belief that you have also fixed milepoint and events. I have found that is not the case. 

Hope this helps.

[KyleGonterwitz]

Esri once shared a script to classify self-intersecting polylines, what it does is work out the type of self intersect types like "Branch, Alpha, Barbell, Lollipop" etc.  It was really helpful for us in cleaning up our network for Roads and Highways.  I'm sure it must be available somehow, Clive do you know what I'm talking about?

[CliveReece]

Yes Kyle.  Thanks for the positive feedback.  Our Esri Transportation Practice has developed a custom in-house tool for classifying self-intersecting routes.  We use it in helping our Roads and Highways customers sift through their data for route shapes that are not supported by Roads and Highways.  The results from the tool are a table summarizing the types of self-intersections found and a feature class of the input routes with a field added that contains the type of self-intersection.  We find the tool is extremely helpful in sorting out and prioritizing Data Reviewer results for self-intersections or polyline closes on self.  

[CliveReece]

Preston, 

Records flagged by Data Reviewer as "Invalid Geometry" may *not* be critical issues.  I would categorize them as "Requires Review". 

Case in point that is easy to test:  Create a lollipop shaped polyline as a single feature.  Run that feature through the Data Reviewer test, Check Invalid Geometry.  The record gets flagged as "Invalid Geometry: NotSimple".  The issue is that the feature intersects itself.  Nevertheless, lollipop features are supported by Roads and Highways for calibrating routes.  In this case, the "invalid geometry" result by Data Reviewer is a false positive, non-issue for Roads and Highways.

(aside:  some lollipop shapes will be problematic if they are multipart features with parts out of order with the calibration direction)

Again, I'm not saying all "invalid geometry" records are fine.  Many will be proven to be issues after looking into why Data Reviewer has flagged them.  And that is my advice, use "Invalid Geometry" as a call for further investigation.  You are going to cause yourself unnecessary anxiety about your data if you see a bunch of "invalid geometry" records and assume they are all critical issues.