Why do we still use point numbers, at least as our main point attribute?
Can't we, like in GIS if I'm not mistaken, have point objects that just have attribute description(s) and have the point object be differentiated by the coordinates/elevation? For that matter so what if we have two point objects's of the same coordinates on top of each other when you have additional attributes to differentiate them such as observation time?
Can any modern CAD software like C3D or Carlson handle data like this? Are there database limitations on this?
Tom
I'm just a construction surveyor, but I would think it might be a little difficult to inverse between descriptions. Maybe I read your post wrong?
-JD-
Construction surveying and field computations in general are an interesting challenge I didn't think of...I was thinking more for general location. But aren't point numbers a semi-new invention in the age of computing? When I look through old field books I don't see point numbers...I see a measurement to a description...It may be BLD1 or BLD2 or MON but not a number.
You could inverse between (for column lines) COL-A and COL-B or for intersections COL-A/1 and COL-A/2. For monuments you could inverse between Mon1 and Mon2. You could inverse control by attribute CP-5 to CP-6.
Tom
If I understand your question correctly, I think it's because point numbers are the unique "handle" to reach the co-ordinate values.
It's a lot easier to draw a line between pt 100 and 101 than between
60013.2451,3232.4256 and point 70103.2215,4220.8556
Point numbers allow you to check data easier and to double check data by checking the "batch sequences" most crews use to track their data.
If you don't want point numbers, don't check that box before importing your points.
But with field to finish the number has nothing to do with the linework...you can start BLD+7 and then shoot two more BLD shots then turn and shoot a POLE shot then come back and code a BLD CLO shot to finish the building off with no point numbers.
Or you can draw a line from CP-1 to CP-2 and from MON1 to MON2...or in a graphical environment you can just pick the point objects from the screen you want to draw a line/measure from.
I think I understand what you mean but can you give an example of how point number rather than a description or other identifier can't be used for the same checks?
I think there are more reasons for them than against them. :-S
Main reason I use point numbers has to do with my field notes. I want a positive identification linking a set of observations/measurements in the field book to whatever survey/drawing I'm working on. I want to know at a glance where that coordinate came from. That being said, I frequently do use alpha numerical point numbers to differentiate GPS corrected data from total station data. For example if I have a set survey nail that was traversed in, I'll give it point number 40. A post processed static session on the same point would be point number 40_GPS, an OPUS solution, 40_OPUS. The software doesn't care what I call it. Mostly I just want to keep the snap, crackle, pop in my head to a minimum.
I don't know, just a thought I had for conversation. I see the reasons but I just wonder if it needs to be the primary method of identifying what a point is when, at least to me, it seems that the description would be more important. When I look at a point I say point 1006 = a building corner. Why not just look at it and say "building corner" (which is the more important information). It's like using number codes in the field...I get why it was useful when we had older data collectors or instruments with no keyboards but these days it just doesn't make much sense.
Anyways...just food for thought for the weekend. 😀
"You gotta be kidding me", suggesting something other than point nembers!
In your example of having multiple points in close proximity being able to select a position by a point number would remove the possibility of selecting the wrong but close in proximity position by graphical methods.
Plus using a number keypad is very fast for cogo operations. Much faster than zooming/panning about and clicking on each point.
Just my thoughts.
> Main reason I use point numbers has to do with my field notes. I want a positive identification linking a set of observations/measurements in the field book to whatever survey/drawing I'm working on. I want to know at a glance where that coordinate came from. That being said, I frequently do use alpha numerical point numbers to differentiate GPS corrected data from total station data. For example if I have a set survey nail that was traversed in, I'll give it point number 40. A post processed static session on the same point would be point number 40_GPS, an OPUS solution, 40_OPUS. The software doesn't care what I call it. Mostly I just want to keep the snap, crackle, pop in my head to a minimum.
That's the beginning of what I am hitting on...there is more data to the main point identifier than just a number. And I get numbering certain sequences of points (control for example) but for other features...who cares what the point number of that one in 500 trees is? It's a 18" Maple, it has coordinates of XXXX, XXXX, and was observed at 10:52:26 on 3/19/14. The data of from where it was observed from is maintained in a the raw file...From Control Point (CP) J6-PK (Leg J, Point 6, PK Nail) backsighting J5-X (Leg J, Point 5, X-Cut). From that info you have all the same data you have when using point numbers...probably more.
When you export Leica (and probably others) proprietary project data, all of this info is there.
Tom
CAD guy: Hey field guy can you describe the method you used to locate "building corner"
Field guy: which building corner?
CAD guy: you know... The one by the door and sidewalk shots...
You get my point... A number is far more unique than a description.
Thank you, makes sense from that perspective.
A point ID remains the most simple way to identify an observation and give it a unique status.
Can you imagine having your driver license ID recorded in the database as "third blond guy on the second street past 2 houses down from the fire hydrant used sometimes by the dog of Mrs. Doolittle"
I am guessing the data sets of your daily jobs are very small.
> Why do we still use point numbers, at least as our main point attribute?
>
> Can't we, like in GIS if I'm not mistaken, have point objects that just have attribute description(s) and have the point object be differentiated by the coordinates/elevation? For that matter so what if we have two point objects's of the same coordinates on top of each other when you have additional attributes to differentiate them such as observation time?
>
> Can any modern CAD software like C3D or Carlson handle data like this? Are there database limitations on this?
>
> Tom
In a database sense, the point number is the primary key. It is unique to each record and therefore eliminates mistaking one for the other, etc. For working with measured data I think this is the best system I have seen anyways as it requires no additional thought processes to make it unique and can easily be searched (ie. incremental numbering). Descriptions of features may be very similar and require time-consuming effort to make unique. A system is what makes data collection efficient, isn't it?
Having said that, for a GIS database used to compile a large collection of control points for example, the point numbers become unnecessary almost. You have no connection to the number and it means nothing, until you extract it for field use or something like that, of course at that point (!) a number can be arbitrarily assigned.
In summary, my take is that it facilitates a system of data collection in a way that removes thought from assigning the primary key attribute to measurement records.
Jacob
In my experience with GIS every feature has a unique identifier, a point number in other words. We just don't see them in our normal interaction with the software.
> In my experience with GIS every feature has a unique identifier, a point number in other words. We just don't see them in our normal interaction with the software.
Yeah, the feature ID is invisible to the user unless the user actually goes looking for it. Like you said it is essentially a point number, so feature ID = point ID = point number, no matter how the user really interacts with the data set I suppose.
Good point, so maybe just refer to the point number as a feature ID and the whole thing is self explanatory? Seems like most data is made unique in almost every type of software database by some sort of incremental identifier, whether the user cares or is able to relate to it or not.
Actually I can see a time where point numbers aren't the first thing you gravitate to. As data becomes more 3D and attribute rich (think Trimble V10, Laser Scanning) I use point numbers for control to maintain my sanity, but the data gathered goes straight from raw sensing data (photo, LiDAR, etc.) to CAD geometry.
If you have a well-oiled coding system and can make it hum like a machine the CAD is 90% finished before you get it in the door these days, anyway.
With the advent of L5 adjustment software IN THE DATA COLLECTOR should be able to automatically build a least-squares network real-time. Proximity to previous observations would automatically alert the system to redundant measurements and the RMS for the observation would determine it's weighting. This would all be code-driven so that it's not trying to average a bottom corner of sidewalk with a top corner and see a height bust, control classification codes would enable the automation.
This obviously gives no thought to stake-out, give me my point numbers there...when I'm not staking straight from a CAD file, which I do more and more.
In the Trimble system your point number is already secondary information - eveything is getting an observation ID behind the scenes.
All that said, I definitely don't see enough surveyors buying in to the workflow to see it happen any time soon, I was merely positing that it is possible in a workflow sense.
