pie in the sky?

[camelfilter]

Hi All,
This is probably ridiculous but is there a system that can project a 1 metre interval grid on to a concrete floor from a height of 15 metres over an area 100m by 200m with an margin of error of less than 5mm.

I'll get what I deserve no doubt but thanks anyway.

[danielbriggs]

Hi,
First off... welcome to PL!
Where are you based, and what is your background?

Ok, as it stands... no, not really. If you were to project vertically down from 15m, the scan angles required are extreme (~160deg at that distance!), not practical
That is not to say what you want can not be achieved!

If you were to project "horizontally" and correct the resultant image with software tweaks, to map the rectangle with super-excessive vertical keystone, you might get away with it.
How big an area is available? Moving the projector back from the scanned area would let you get away with making the scanners scan less wide, for the same image size.

If needed, a lens could be added after the scanners to magnify the scan to achieve the required size; although beam size and accuracy would also scale.

To add to the headache, a grid of that density would be very demanding indeed to scan at an acceptable rate! Another possible solution, should 'conventional' means not be up to it, would be to have two scan sets; one scanning the vertical lines, and one scanning the horizontal lines.

In addition... the projector(s) must be mounted very very stably to even approach anything near the accuracy you require. Can you budge on the positional tolerance? At 200m, the best resolution you can hope for is 12.5mm, and that is the software resolution, not taking into account the actual positional accuracy of the scan set.

If the installation is going to be fixed... it might be worth investigating using 300 small laser modules with line generating optics attached (that would be the easy part!)... accurately positioning and aligning such would be tedious and expensive for appropriate mounting. (The more I think about it... this approach would be the best!)


Finally; I have to ask: what is the application? Some sort of high-tech envrionmental survey grid marking system? Or a human size gigantic game of connect-4?

Best Regards,
Dan

[-bart-]

Diffaction limit would prohibit a line thinner than 5mm at 100m distance at 10degrees incidence (at a concrete floor).

*possible* IF such a system should consist of multiple projectors.

I would not go the closed loop-scanning way, only the slightest drift will ruin your accuracy.
Custom HOE with perfect pincushion correction maybe.

[camelfilter]

Thank you so much for taking the time to reply guys. I am completely without knowledge when it comes to this field. I have a big problem.The project is to drill 14000 holes with a tolerance of 5mm into a concrete floor in 20 hours. Normally we would use tape measures & chalk lines to mark the points but there is no time. If we could work from a grid it might be acheivable. From your answers it's safe to assume that a rig like this would have to be custom built at significant cost and would be incredibly difficult to position. Thanks again for the info.

[danielbriggs]

Whoa, that's a hole every 5 seconds!
How about making some jigs; and use a cheap laser of low power with a line generating optic for edge alignment of the jigs? It should get you close!

Best of luck

[camelfilter]

I have twenty guys on this Dan. Problem is the positioning of the holes is completely random. There is no uniformity so we can't really rig anything.

[Jem]

Is it me?... Quite possibly it is, as i'm an old fart currently writing from the retirement home for the terminally bewildered.

First you state...

...with an margin of error of less than 5mm.

Then you write

...positioning of the holes is completely random

Any chance of a bit of clarification so that my ageing brain can come up with a solution

[allthatwhichis]

You might be able to use a Discoscan lense. It can get you a scan angle of 160 degrees +, but I am not sure about how "square" the image will be. When do you need to do this? I can try to take some measurements this weekend with the Discoscan I have at home.

[camelfilter]

Okay,so the positioning of the holes appears completely random however the positioning of each hole is critical. Each hole will be threaded ready for a structure (I can't even begin to describe the shape of it) that will arrive on site in sections ready to be bolted to the floor. I've been set a tolerance of no more than 5mm for each hole. In a parallel universe we could have something attached to the roof that would beam down 14000 dots onto the floor simultaneously in the correct positions but that ain't feasible hence my question about the grid.

[norty303]

I'm confused by the need for a grid, if the positions are random (or at least not uniform)
When you first started, I thought you would be drilling at each of the intersections, but it appears thats not what you want.