Azimuth Technology for Cave Surveying

kdxn

kdxn

New member
Given the performance of the Disto D8, up to 200m range (to reflective tape) and +/-0.1degree inclination, the limiting factor for cave surveying would appear to be magnetic bearing accuracy.

I would be interested to know whether anyone has experimented with modern gyro north seeking technology and/or yaw sensors to determine absolute azimuth or the angle turned at a station between backsight and foresight.

Needs to be easy to use, small, low price(hopefully) and caveable.

Would appreciate any posts about experiments here and abroad.
 
graham

graham

New member
Experience with a DistoX would suggest that the current limiting factor with hand-held kit is the ability of the user to hold it steady.
 
TheBitterEnd

TheBitterEnd

Well-known member
I have no recent experience but I did do some gyro mine surveying at college. Apart from the kit being reasonably heavy and fairly delicate, the issue was the length of time it took to take a reading. The gyro precesses and it is necessary to take readings over and hour or so (IIRC) to find the end points of the precession. Modern instruments may automate the reading process but I doubt anything can be done to reduce the time required.

That said, I guess it depends on the accuracy you are looking for. Again IIRC we were going for much better than one minute of arc (tens of seconds), it may be possible to operate quicker for lower accuracy.

I guess if you had a long and fairly straight passage somewhere in the cave you could put in a base line against which you could correct magnetic readings but I wouldn't have thought it was practical for lots of short sights.

I have very little cave surveying knowledge, but I would have thought that there were other techniques that would improve angular accuracy.
 
kdxn

kdxn

New member
The old style gyro technology such as the GAK1 which bolted onto a Wild optical theodolite has been used a lot historically for mine surveying.

Some mine surveyors have used tactical grade inertial sensors but these are very expensive and usually big and heavy. Okay if you are driving along a tunnel in a truck.

Granted a Disto needs to be held steady and shots need preferably to be between stations but that is not always possible.

A yaw sensor would require an angle to be turned between backsight and foresight so has some downside there. There are some small low cost sensors.

North seeking gyro chips are available but the drift rate precludes a lot of them unless this could be adequately compensated for, perhaps with the right observation procedure.

The Disto D8 has a small daughter board inside for the tilt sensor. It would be great if this daughter board could be modified to include a compass/angle sensor.

Might be a useful Masters project for someone ?
 
kdxn

kdxn

New member
Thanks for the link. There are quite a few options for small nav boards in the robotics community.

My interest was to go beyond this and try something different, hence the request to see what others may have been up to.

There are some boards available where the high rate data can be integrated to provide a plot of where you have been, a surveying solution in your pocket. Accuracy is very dependent upon sensor price, regular calibration and post processing parameters.

I have a low drift gyro chip option that can potentially provide an accurate absolute true north periodically through the cave and turned angles at every station. Need to bench test it against something a lot more accurate. Observation procedures would be a bit more cumbersome than the current typical cave surveying setup but potentially more accurate and robust.
 
graham

graham

New member
kdxn said:
Observation procedures would be a bit more cumbersome than the current typical cave surveying setup but potentially more accurate and robust.
Click to expand...

Out of interest, what sort of degree of accuracy are you aiming for? We are currently getting figures like this:

REL-ERR ABS-ERR TOTAL-L STS X-ERROR Y-ERROR Z-ERROR

0.27%    0.2m  80.1m  15  -0.1m    0.1m  -0.1m

0.22%    0.1m  35.4m  9  -0.0m  -0.0m    0.1m

0.16%    0.1m  37.5m  6  -0.0m    0.0m    0.0m

Those are the last three on the log file of the current project.
 
kdxn

kdxn

New member
Some good figures.
I note the small distance and small number of legs. For bigger stuff where you can use the D8 near it's maximum length, I would like to see an azimuth accuracy of better than 0.5degree per leg.
 
graham

graham

New member
kdxn said:
Some good figures.
I note the small distance and small number of legs. For bigger stuff where you can use the D8 near it's maximum length, I would like to see an azimuth accuracy of better than 0.5degree per leg.
Click to expand...

I don't understand the maths, but I am reliably informed that it is better to have more small legs than fewer long ones. Those particular loops were just the last ones on the list, though representative of a cave that has an insane number of interconnections. The job has taken far longer than we anticipated when we started.

If you want an accuracy of better than 0.5 deg on legs of 200 m then you are going to need to mount your instruments - and your target - on tripods. That takes me back to the discussions about surveying in the early 70s when hand-held Suunto instruments first became widely used.

I think you will find that most cave surveyors will accept a small loss of accuracy for a vast saving in time and convenience.

And anyway there are few caves in the UK, few in Europe really, where 200 m legs are going to be possible.
 
kdxn

kdxn

New member
Its all down to stats. Lots of short legs will average out random errors but will increase biases from station setups. Your recent loops would suggest that you have some good procedures for station setup and observations.

My interest goes beyond the handheld. I would like even more azimuth accuracy to help with laser scan registration.
 
cavermark

cavermark

New member
Any surveying that involves passage detail drawing is easier and more representative with shorter legs.

I guessing Kdxn is looking at line surveys to tie in Laser Scans to surface benchmarks, so 200m legs are fine.
 
kdxn

kdxn

New member
A big issue with laser scanning caves is how to transfer control. Normal technique is a traditional traverse but this requires station setups at sub-optimal scanning locations which all require time. Alternatively temporary targets are setup visible from adjacent scanning locations with good orthogonality - angle fixing ability - but again these all take time.

Having an independent accurate azimuth determination on the scanner will reduce the uncertainty when matching overlapping scans especially when using a coarse scanner. If there is sufficient overlap in the roof then you can effectively traverse via the overlapping roof observations and avoid the time and complication of small traverse legs down at the floor.
 
A

AdM Michael

New member
Depends what technology you're using:

Compass (magnetic): short legs as a wrong reading will then only cause a small parallel shift of the following legs
Theodolite/Total Station: as long legs as possible as a wrong reading has an influence on the bearings of all following legs

Just have a look at the covariance propagation law.
 
kdxn

kdxn

New member
Ahhhh, the biggest passage, Deer Cave or not Deer Cave, that is the question.

Which can not be answered until someone lets me or someone else with a long range 3D laser loose in HSD. Not a communicable disease, but something reputedly big in Vietnam........is it bigger than Deer Cave, does the passage height exceed the new passage height in Deer Cave, does the passage cross section exceed the biggest cross section area of Deer Cave ?

Would be nice to have some answers at HE2012...........
 
You must log in or register to reply here.
Top