I love it when a plan comes together!

Getting a precise plan of a building like this needs a good control network: I went for a classic  loop of stations to fix the exterior footprint with a spine though the building to hang resections from.
With this project I needed a fair number of photo control points for the interior facades and  discovered mixing data types in a single network calculation may not be a good idea…I shot in my photo control points as reflectoreless obs (2 rounds per target) on the same network; when I ran the calc it was not a pretty sight. The wailing and gnashing of teeth began as I saw the calculated displacements: all that red in the table is not what I expected! How could this have gone so wrong?
I slept on the problem and then re-ran the calc WITHOUT the photo control points:
And the results were much closer to my expected values..whew! So what went wrong?
TheoNet Adjust works by distributing the error across all stations, there is a bias in the distribution to stations with observation sets in BOTH directions. Unfortunately in this case the large number (100 or so observed from 4 traverse stations) of photo-control stations was sufficient to distribute a disproportionate error across the 17 stations in my traverse loop and corrupted the overall result.
I know my network was not badly observed as each shot was checked at observation and orientations were retaken if there was any ‘brown data’ displayed in the table. Dud shots were weeded out and I updated my CAD plot on completion of each facade with the ‘view’ command to get a visual on where the shots were placed. My worst recorded error was 20mm in z so I was expecting adjusted positions to have corrections of less than 20mm. Once the photo-control points were dropped from the station list this was indeed the case.
Once the control is in place getting the detail to fit is easy The external trace allows me to fit the measured drawing details, I scan them and insert them into the DWG to make the work easy;Plotting the site notes is a case of building up the geometry by offset from the TST trace. They take a while but it pays to get the detail right.
The plan took around 8 days including site photography and sketches. CAD work was another 2 days to get the details in.  Using TheoNetAdjust was a big help  for 2 reasons: 1. I was able to keep a running check on the precision of the control as I was measuring it and 2. I am able to move all data by station post adjust: I was able to trace all I could catch from each station and be confident I could refit it post adjustment if I needed to.
The geometry of this plan reveals some interesting historical details and getting the plan to record this correctly  can only be done reliably with a proven adjusted network.
Learning the behavior of the network algorithm in TheoNetAdjust  was something I did not expect to deal with but I was able to prove my loop in the end!
Happy surveys,

B

About billboyheritagesurvey

Heritage worker
Gallery | This entry was posted in Survey Practice and tagged , , , , . Bookmark the permalink.

2 Responses to I love it when a plan comes together!

  1. Jonathan Turner says:

    Just Brilliant Will!

    Very interesting read… What results!

    Jonathan

  2. Pingback: traversing | Billboyheritagesurvey's Blog

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s