Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Tilt compensation change from APC to pole tip is hazardous to data integrity #499

Open
gdt opened this issue Nov 7, 2024 · 0 comments
Open

Comments

@gdt
Copy link

gdt commented Nov 7, 2024

(I don't have a torch, but I know someone who is thinking about it.)

Currently, with the non-torch units without tilt compensation, the workflow is:

  • unit reports position of APC
  • user levels the pole
  • data collector is configured with APC-tip offset (= APC-ARP offset plus ARP-tip offset!!)

and this is ok, except you don't have tilt compensation.

With the torch, as I understand it from reading the docs, when you first get an RTK fix, the unit is outputting the APC position. Then, after tilt compensation, it's outputting pole tip. But if it pops out of tilt mode, perhaps due to momentary loss of RTK, it can go back to APC. This means one's data is suspect, basically.

I suggest (assuming I'm not confused):

  • user manual explains that "pole height" is APC-ARP plus ARP-tip, and SF or somebody experiments with and documents the APC-ARP average value
  • user configures APC-tip as happens now
  • when not in tilt compensation mode, the APC-tip offset is still subtracted, so one gets pole tip assuming pole is vertical
  • if tilt mode is lost, then output just stops, until either regained, or user explicitly turns off tilt mode.

The point is that incorrect data should not be recorded. Already data collectors have to record the error in the M field of POINTZM, or refrain from recording if high, but tilt/not is an additional dimension.

@gdt gdt changed the title Tilt compensation change from RX to pole tip is hazardous to data integrity Tilt compensation change from APC to pole tip is hazardous to data integrity Nov 7, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant