Materials / Texture option

[digitalcarbon]

if we do this we could eliminate most annotating 

 

 

[Markvl]

Send it to the wishlist.

[bcd]

I'm wondering what you don't like about sheets. I agree with your argument for materials, properties and metadata to be embedded and visual on the model, I'm just not convinced that the best way to communicate the design is always via a 3D model. 

I mean for example the shop wants to fabricate the frame for your ductwork. Do they rotate & pan the model to read off and write down the various , materials, lengths, holes angles, welds, connections etc repeatedly during fabrication or has that been tabulated graphically somewhere & pinned to the wall. Sure,  cutlists can be generated automatically but 2d views are not obsolete- often they are a much more meaningful and pure way to convey information.

i could see a hybrid of what your proposing where 2d data-rich views are automatically generated from a data-rich 3D model. The 2d could be available in orthographic projection either on an automatically generated sheet or in a saved view. One click on the model to a 2d saved view or sheet with data. Thoughts?

Edited February 5, 2017 by bcd

[Matt Overton]

Might take a bit of tweaking to get the density of tags right but I can see how we would make use of this.

Would be useful if class over rides allow it to be turned off on selected viewports.

[digitalcarbon]

bdc, what I'm finding is that i have to stop modeling (and i have a ton of modeling work) to make 2d sheets. what i would like is to do is share design development, concept info as 3d only but  in the end invite all people of the project to harvest any information they need out of the 3d model by way of a web based VW free option (that pdfs directly to bluebeam rev).  what i am finding now is that I'm busy making ice-cream and do not have time to wait on customers making cones, fudge sundays, banana splits.  if i do then i run out of ice-cream.  sure i could hire some people and i may do that but i need to keep making ice cream.

 

[Matt Overton]

I wonder if the objections to "drawings" would be a lot less if the double handling of information the computer already knows was lessened to significant extent and we could mostly concentrate on telling the story. 

 

[digitalcarbon]

Matt, bingo!  what I found with the last large job (a 110,000 ton/year composting facility) was that the designer needed to tell a story of how this unusual system worked.  i did his modeling and made videos to replace the "napkin" sketches that most designers use to convey ideas. the model was accurate & i tried to set up a 3d export relationship in the beginning of the project.  that did not work due to the other team members having only a limited experience with 3d work.  as the project went on and more sheets were made by the 2d thinkers, i & the designer lost more & more control of the project.  large sets of 2d sheets were produced & we started to see deviations from the model & asked everyone to match the 2d work to the model but that was not well received.  we were suspended from the project for about 6 months because everyone thought the modeling work was done.  then we were hurriedly asked to come back and figure out some situations by way of modeling.  while i was modeling i found a few major problems (concrete wall 9ft out of place,  no head room for a man-way access, 40ft container not tall enough to fit equipment, etc)  now everyone was more willing to pay attention to the model. 

 

 i did produce a few 2d sheets when the various engineers asked for them, but this was slow (rendering hidden line took forever) and had to resort to Open GL rendered view ports.  this was actually a +  because it allowed me to label less because people could see the textures of things i modeled (tried to get things to look like real life) and understand what things were with out me labeling everything.

 

the lessons learned are:

3d modeling is the best way to control a design (one source)

not everyone can work AND navigate in 3d (there is a learning curve no mater what software you use)

3d modeling requires state of the art hardware & software (offices just do not have the ability to keep current,  some could not even deal with Skype)

3d export was a job in itself (with all these different people working on the project exporting/importing became a 1 day a week operation)

 

what i would do next time.

a small team of 3d modelers, broken up as follows:

3d assemblers (builds the model from any existing, adds products, dtm, fits to survey grid etc)

3d symbol makers (all products to be in the model)

3d to 2d export dwg makers (these take the model and extract plans, elevations etc for the engineers to place in their drawings)

 

final wish:

VW needs to be web based, with my 3d team paying for our seats.

then i can invite anyone to the model to harvest anything they want out of it. (for free) like a self-serve buffet

[digitalcarbon]

another lesson learned was,

the few subs that did have a guy who worked in 3d did not work as clean as i did.  

what i mean is that he modeled just enough to make a drawing for the guy in the shop to fabricate and any discrepancies they would iron out in the field.

 

so when he exported a 3d model to me it did not fit.

Edited February 7, 2017 by digitalmechanics

[digitalcarbon]

a bit more on what i would do next time.

 

i would have the 3d assembler in a type of "face time" set up with a person who is in the field.  

 

as things are being built the 3d assembler can see what is happening in realtime.

 

then the assembler can have the model pulled up on an second screen and be verifying the work done. just as a double check.  

 

then if something needs to be moved in the field (conduit for example) the modeler can just adjust the model on the spot.

 

the way it works now is people are running around the site taking video & pictures then dumping them on the 3d modeler at a later time.  

 

a flying drone, controlled by the modeler,  would be the best but the tech is not present yet as these cannot fly for any usable length of time (8hrs)

 

[Diego - Resuelvector]

4 minutes ago, digitalmechanics said:

 

a flying drone, controlled by the modeler,  would be the best but the tech is not present yet as these cannot fly for any usable length of time (8hrs)

 

It's a great idea, one could do a fast checkup how the works in the field are going on.