One man/Robotic Construction Stake out......

I have a question for my fellow surveyors out there who are in the same boat as I am. I am a union surveyor and I work in the heavy-highway industry. Mainly doing bridges, roadway construciton and some foundation/building work. I have been working for the last four years with only one companion...my Leica TPS 1200 Robotic. I have often found myself in a situation where I have to lay something out, for example column lines for anchor bolts...and I know how I would do it if I had an instrument man and a Topcon 303 B...I would set the gun on a column line, take my backsight and turn 90's till the cows came home....but now that its just me and my 1 second robotic instrument, how should I proceed? Even If i did set it up on every column line where I needed to set a stk and turn 90s with the Leica, I am still holding the prism and measuring like I would if i did a radial stake out with a conventional gun.

In the past I would always have the gun set on some offset to the column lines no matter what. I would never just have a random fly point with no tie to the building lines...but from that point I would stake any other offset points radially. I am obviously performing random check distances between all of my stakes, and making sure I always check back into a third or fourth control point every time.

I am curious to learn how someone else who is in my position might approach this. Especially in instances where my tolerences are small.

Thanks,

--caleb

You'd just waste time moving setups and turning 90s.  In fact, using your robot for radial stakeout from solid control points with short foresight distances can even get you better results than moving your setup around on layout points, taking short backsights, and turning 90s.  Just make sure to always pull adequate checks between layout points.

I think that you should turn your 90's, frankly. What you describe is one of those cases where the robotic total station is not the best tool for the job.  I've sometimes thought of picking up a T1 and a steel tape exclusively to use on that sort of work. If I did more of it, I would.

If that's not a option for you I'd at least put the prism on the bottom of the rod instead of the top to minimize the out of center-ism.  And check, check, check as Richard says.

I would establish a very solid control network around the project, probably via traversing and levelling.
Then I'd setup close to the area in question and resect in from the surrounding control using a miniprism - perhaps a leica mini360 set as low as practical.
Then I'd setout all the marks in the area using a mini prism, set low using radiations.
Then once complete I'd check my control again. 

Obviously you should use long backsights, check your work with tape, etc.

I'd take one solid setup that's well constrained over lots of setups chasing column lines anyday. 

The idea of setting up on line, etc is useful, until you end up finding a job that's all curves, or where you can't setup on the line. Then you have no choice but to radiate your control without the experience gained from utilising that technique elsewhere.

If in doubt look at the errors. Look at the options and see which gives a better result.

Oh yeah, don't skimp on decent equipment, especially mini prisms if you use this technique.

Radial stakeout constrained from multiple, stable, back sights that surround the work, can be much more accurate than the "tried and true" setting up on a column line and cranking 90s.
 
As mentioned, you need very solid control network around the project.  We’ve have shown on industrial metrology jobs that this control can be traversed using laser trackers.  This type of control network is much more accurate than conventional optical metrology methods.
 
I believe, people still like the column line and cranking 90s method because it brushes over a good deal of error propagation issues, e.g., monument stability, tripod stability, instrument centering errors, targeting errors, etc.
 
People typically use levels for vertical control for many of the same reasons; it brushes over a good deal of error propagation issues, e.g., equipment understanding, tripod stability, instrument centering errors, targeting errors, etc.
 
How much does the monumentation move during the course of the day/season/year?
How do I constrain or model the monumentation movement?
How much does the tripod move during the course of the setup?
How do I constrain or model the tripod movement?
How can I better value the center of my total station?
How can I better target the item I am surveying or laying out?
Etc., etc., etc.
 
As you better understand your equipment Datasheet, the better you will understand the performance and proper use of your equipment.
 
I always recommend checking your work, especially when first implementing new procedures. Eventually your skills and confidence will increase to the point you will not feel as compelled to check “everything”.  Checking your procedure will become more commonplace.
 
Radial stakeout, when properly done, can be faster, more accurate, more precise, safer, less expensive and easier than the older methods.
  

Thank you for all of the feedback....it sounds like I have been on the right track at least. I always use a small peanut prism when the situation allows it for doing radial stake-out,  and I always use fixed tribrachs for backsights and foresights for my traverses.

I am a little leery of the TCP 1201’s ability to accurately perform high precision stakeout in robotic mode.  And that may only be because I dont completely understand the ATR feature.  The 1201 will achieve excellent precision and closures when measuring fixed points when ATR is turned on.  I don’t believe that ATR can be turned on when staking a moving prism.  If we fail to turn ATR on when running control traverses, then the closures are usually barely acceptable.  10 point traverse around 10 acres with ATR on and in robotic mode will almost always achieve 1:50,000+ closure.  Same traverse with ATR off will usually be less than 1:20,000.  It's a mystery to me because the cross hairs never point at the actual center of the prism with ATR on or off.

We will occasionally perform high precision staking with our 1201 and we do it radially but turn off the robotics mode and work conventionally.  I want a good I-man lining the bottom of the rod up with the crosshair.   We always do a limited amount of checking by occupying various staked points and turning 90’s along with measuring diagonals. Too much liability to be wondering if it is accurate or not.

If someone could further explain the ATR feature on the Leica TCP 1200 series, that would be great.

Mitch, can you please summarise for me when you are using ATR, Lock and manual?

When you say ATR off, are you meaning in Lock mode or manual?

FWIW, when staking out points as described in this thread so far I'd usually use the ATR over lock, simply because I don't want it searching all the time. However, there should be no difference in results whether you use Lock or ATR.

Mitch .. ATR is fine.  Leica would not have introduced it if they thought it was not accurate! ATR has software that computes the offset difference from where the instrument first points to the centre of the target and then applies a correction to the initial readings of horizontal and vertical angles. This was introduced to stop the telescope wasting much time trying to  zero in on the target's absolute centre. You have to forget looking through the eyepiece and noticing the cross hairs are not aligned with the centre of the prism. Believe it or not I use ATR and lock for most jobs. Only manually observing in poor atmospherics  and or with long sights out of range of ATR.

Caleb.. One point missed out by other posters is that when setting out your points you can visibly see on screen the final delta x,y, z from design coordinate. Obviously close to .000 for x and y. (z could be a  difference to design) . So that you are happy that you have set out the point, for peace of mind or a potential future dispute (especially if a set out point is knocked by an unknown party and not reported) I then record that set out point data as if I were picking up a point with a unique number. ie it may be the original point number with a suffix like F ( for field set out point ). Then reduce the set out file into the data file so that you can then check that you have orientated correctly and then generate a point comparison file by nearest points eg set nearest point range say 1 cm. That way you can insure that you have not read for example a 6 or an 8 as a 0 and that you have proof that a third party has  knocked , moved  and or replaced a set out point.

I remember  when at university  playing around with error theory of comparing  setting out by conventional grid and radiation, (of course 40 plus years ago we did not have  today's technology), but I recall radiation would be more accurate even using the then limited conventional surveying equipment

RADU.

I worked on a job that had over 100 piers and bolt layouts.  We could not put in good centerline control which made us radial stake everything.  I used 4 control points for everything that required a layout.  I also had 4 extra control points that I used as checks and when there was a problem I could isolate it.  I also sued the least amount of rod as possible and had 2 level bubbles with me al the time. I was not a very big fan of doing it that way but got more comfortable as the steel went up without any issues.  I used a Trimble 5600 for the entire job, including most elevations and had no issues.  Well there was the bobcat driver that hit the form and did not say anything until the concrete was dry.  I guess the thing is to just check, check, and recheck.  

As for turning 90's.  We do it all the time, and never have problems.  BUT, we go a little farther than most I think...not sure, I just think we do.  I have set a number of buildings and equipment turning 90's and when done right, always works out.

However, the Laser Tracker is the way to go, if you have a 100g's or more.  If you can measure as close as .001" then you have the upper hand.  However, I have heard they are not good in sunlight, dust and humidity.  Falk can tell you more about that.  I also think they are short ranged, less than 200'.  We have worked around them before and it is a nice product, however alot of carpenters still like 90 lines so they can string a line.

My thoughts.

Sprint Car:
As for turning 90's.  We do it all the time, and never have problems.  BUT, we go a little farther than most I think...not sure, I just think we do.  I have set a number of buildings and equipment turning 90's and when done right, always works out.

However, the Laser Tracker is the way to go, if you have a 100g's or more.  If you can measure as close as .001" then you have the upper hand.

You typically don't get yourself into problems turning 90's, unless you backsight short and foresight long.  But if you have decent modern equipment, you also aren't gaining anything, and are basically just getting the job done much more slowly than necessary.

As for measuring to .001" in the field...  Yes, some jobs have less tolerance for error than others.  But that can be pushed to insanity...  If you really need to care about .001", then your entire work environment should be completely enclosed and environmentally-controlled.  Otherwise, you are just fooling yourself.

“…However, I have heard they are not good in sunlight, dust and humidity…”

Refraction and Diffraction affect all optical/laser equipment. Sunlight can also cause temperature issues with your tripod and targeting systems which normally are not modeled for typical construction surveying. Outdoor Metrology typically will need to be done at night to work too 0.001” levels.
 


“…I also think they are short ranged, less than 200…”

Tighter tolerances are typically also a function of distance.  A Laser Tracker is typically overkill for anchor bolts/can as builts. We prefer to use our Leica TDA5005 or Trimble S8 for anchor bolts/can as builts.
 


“…You typically don't get yourself into problems turning 90's…”

This is true if you know how to take the “split” out of your T-2, T-1 etc.. 

It all depends upon who your final client is, or who is going to use those points.  I was always a firm beliver in the well done radial stakout untill form contractors started using our stakes to "turn their 90's and check our distances". Funny thing was, no matter how careful we were and how recently calibrated our Leca 1100 gear was, when we went back to check our own 90's....they did not check!

I have gone through this exercise multiple tiems on all manner of projects big and small and in the end I determined that when it comes to squares and rectangles, nothing beats turning ninties.  So our mantra now is.. sight long and shoot distances short.

We come off our well shot and well balanced traverse (which we have a very warm and fuzzy feeling for) and shoot the two longest points of one long leg (assuming a typical column line building site).  We set the tack on the first point give it a new point number and only "mark" the second one. Then we come down to the first "new" point and shoot the "correct" horizontal distance through the mark of the second point. That is where we set the tack and record another new point number.  This is all done with prisims on calibrated tribrachs and on tripods.  Once we have the 2 REALLY good points we start to turn 90's and shoot the correct distances. We close the "box" and only then do we set intermediate points again sighting long and shooting short with peanut prisims or the Leica reflective cards sighting right on top of the hub.

Even doing all of this, we still find a corner that baffles us and we need to tweak a couple of times.

We are usually setting this kind of control for projects that another subcontractor is going to use.  Now the form guys come out and the first thing they do is say "everything is out 3/4 of and inch". This is to establish the fact to the prime contector (our client) that the surveyors are wrong and we are not liable for their mistakes.

Of course we come right back out for our client and check everything and notice one corner out 0.03' and we fix it and everybody is happy.

Long story short for the original poster; if you are the only one using your control and you are comfortable with it then you can use whatever method works for you. But if you are setting control for contractors (and nowadays it is most likely only 2 points) then you need to think of how they are going to use it.

More than likely they are going to take out their dusty, off market gun from under the other equipment, set up on one point, sight another and turn 90. Low and behold, it does not check!!!

My motto; sight long, shoot short and low and the box is all by itself and will never conform back to the traverse the way we would like.

John

John, But then it is up to the surveyor to elucidate that  1 minute of arc over a distance 100m is equivalent .025m, so over 20 m = 0.005m or  10 " of arc over  100m is 0.004m.. Horses for courses!  Setting bolts  for machinery different from setting pylons.

RADU

Hi there,
I am also a union surveyor working alone now but even when I was a two man crew I used my trimble robot to layout snap lines. The first key is to replace the 20 minute vial on your pole to an 8 minute vial. I use the stake to line feature in the survey pro software and has worked out fine for me for the last 8 or so years.
Hope this helps,

FC Watson L.S.