I'm a surveyor, and our company does a lot of 3D laser scanning of things, all sorts of things. Sometimes we get to scan railway things which can help with modelling. Lately we have been scanning Redfern Station in Sydney, a fairly major one, with 10 above ground platforms (and one un-used dock platform), 2 underground platforms and two "ghost platforms" we have scanned them all. From all this laser scanning we have to make a 3D CAD model of everything. On platforms 4-10 there are 4 A-10/A-11 style buildings, the standard NSWGR brick buildings of the time. I thought when modelling the wrought iron brackets for the awnings I could make a simple cell in CAD, then just apply to them all, after all I have seem people selling laser cut wood ones for station buildings and other people complaining about the brackets in the old AMRI suburban station kits not being quite right. When I applied the brackets from the building on platform 10 to the one on platform 9 I got this problem. The white line is the outline of the bracket from platform 10. The blue/yellow/purple dots is the "point cloud" which is the output of the laser scanner that we use to model all sorts of objects into vector format. These images are from cross-sections with the model and the point cloud showing to emphasize the differences. The horizontal row of points is a water down-pipe. The length is all wrong, but this would have been modified for the clearance of the slightly narrower platform. But it is also worth noting the upper curved part is not the same. The white outline is from a b-spline approximation to the curve of the bracket on platform 10, the bracket on platform 9 is actually an arc, with a fixed radius. The height is also wrong. Other platforms had the curves right but the lengths wrong, like this Finally a picture of the point cloud lining up with the white outlines of the model It turns out that platform 10 was built slightly later than the others, but at a cursory glance they would all appear the same. Without the scanner it would be very difficult to pick this difference up. The moral of the story is don't assume one size fits all if you want real accuracy. I will have a few more interesting scans of Redfern soon, to show the door and window details, but also roof angles. Cheers Tony
Very interesting Tony. I have never noticed the difference in the architecture, especially after catching/changing trains at Redfern for nearly 5 years straight ! Looking forward to seeing more of your scans and the secrets that lie beneath the skin of the NSWGR infrastructure. Cheers, Gary.
As promised I have put my head back in the "cloud" and here are some more scans of Redfern. I have more scans of other Sydney stations, and some Melbourne and outer Melbourne ones, but they will have to wait. Here we have a cross-section of the roof of the building and the awnings. Can of course determine the correct angles for this building from this, as well as the sizes if it was in proper scale which it will not be with this screenshot. Note the difference between the two sides for the awning width. Took the shot with the air con of course! Then we have the side view, these vary a lot, but it has come out nicely in the scan. We have got this in colour too, as we used a camera to colour the point cloud, but I prefer to work in intensity gradients like this. The I have another image that shows the cross fall of the platform surface to allow for drainage, I wonder how many layouts there are out there that don't include this detail on their platforms, when are are talking mm, this becomes important! Cheers Tony
That slope on the platform looks significant - maybe they borrowed Toto's spirit level. Impressive results, I wonder how long before its available to us mere mortals at a price we cant afford. Paul
This is fascinating! What is the company using this data for? And definitely not tooting my own horn.., but you do have Sydney Central? or Newcastle?
We are doing them for Transport for NSW, usually as the first step in any upgrade or major change. We have done lots of stations in Sydney, and yes we have done Central, not Newie though, furthest north we have scanned is Mt Colah and some work around the up sidings in Gosford. The data and more importantly the vector output generally belongs to the client, we can put up small snippets like this, but not the whole thing. Can't share the raw point cloud easily, it's over 300GB worth, might be more. The vector diagrams that we produce in a CAD package are much smaller. It would have been nice to do Newie while it still had tracks, bit late now Cheers Tony
Wow you could almost sell these on like a 1tb portable harddrive and would be amazing to see all stations done across the network both metro and outer areas
Thanks for sharing Dr Tony, it's very interesting! It's now Law in the UK, that any building project in the public sector (government, Hospitals, Train stations, Bus stations, airports, leisure centers, etc) are designed in 3D. To enable this for existing structures, a laser point cloud has to be done, to enable a 3D model to be created. In the same vein, all model builders are getting 3D scans of engines/rolling stock, to get the original 'drawings' into CAD so they can create the most accurate moulds. There is even an online supplier of miniature people, who will scan YOU, so that mini you can be found on your layout!
Would that be ModelU3D in Shropshire? There figures are brilliant! Used them myself - the detail is fantastic
Oh, thanks! I know Alan is stepping up what he can do with a new machine and broadening the scope of his excellent range.
