Even Google Has Its Limits

After I read Scott McEachron’s post on the new functionality today and where they had derived surface information from, I caught him on IM to chat. We’re both big fans of the idea behind the Google Earth interface, but there are some gotchas, and not enough people know about them. Do you? Find out after the jump.

 

 

So if you’re checking out your site in Google Earth, you’re probably in preliminary stage, and just getting a feel for the site. Maybe you’re working on a overall drainage plan, or trying to understand how your site interacts with the rest of the world. In either case, you’re playing at a fairly high level. Maybe something like this picture?

You’re working with high level images, but trying to get some idea of what’s happening, so the PM or developer asks you to pull in some surfaces information. In the past, you’d pull out your USGS Quad sheet, or maybe a DEM. With 2008, you’ll just hit the nifty Import from Google Earth and wind up with a nice contoured surface. Cool.

Then you decide to get a little further into the process, maybe getting some conceptual ideas about how your site will work, so you zoom in a bit in Google, because it’s surely better than that USGS map that hasn’t been updated in 40 years, and it’s way cheaper than sending out a field crew before you know if it will work. So maybe you zoom in closer with GE, hit that magic import, and have something like this. This is obviously a higher resolution of the area of concern (my house when it was dirt for those who care.)

Both of these images have generated surfaces in C3D. I’ve named them GE-High Alt and GE-Low Alt for clarity. When you put them together and turn on the contours, you get something like this next image.

In case you can’t tell from the image, that’s a bleeding mess. The surface data isn’t the same! You see, in order to make GE fast on as many connections as possible, Google only ships out the information it needs to display. This works great for making images as fast as possible, and for creating the 3D views you can access in GE. Here’s the rub: That same limit of detail is passed through the API to Civil 3D! 

What’s this mean to you? Probably not much, but you should be aware of it. Just for illustration, I ran a volume comparison between my two surfaces.

The details are in this next image. As you can see, there are points where the surfaces are as much as seven feet out of sync. I’d say that’s probably within the tolerance of GE in general, but it does mean that if you ever use a GE surface, you really should be exporting out a kmz file so that you can return to the same zoom and view if you ever have to reproduce it.

Finally, don’t design from this stuff. It’s just not that good. We know you shouldn’t do it, you know you shouldn’t do it. Don’t let your client make you do it. Nothing good will come of it.

But man, it does make some great conceptual planning and illustration data. 

Comments are closed.