Take-off issues... How is everyone working around them?

[twiceroadsfool]

Since i started Building modeling a year ago, in the different firms ive been to ive noticed a recurring theme: When we were drafting in 2d, everyone managed to draw things differently. Some people extended lines and trimmed them back, others snapped to points, others typed in coords. It had no effect, as long as the end result was the same. Now, were notcing that people can conjure up different ways to achieve a convincing model, and it swiftly kicks us in the nuts when we start take offs and/or quantities.

For instance:

Edit: Profile. I dont use it, unless im creating a hole in a wall, or an opening, but some people here use it for shortening the length of a wall on the fly. Only... The properties of the wall stay at the original length. Obviously, this is just an in house practice thing that we will resolve by getting everyone on the same page, but that got me thinking...

Wall Types... The wall type that we are using here, for CMU walls, is actually a solid wall, with Hatch on it, called out as CMU. Now, we would obviously not quantify this wall by its volume, but it does get a solid volume in its properties. Is there a way to make a wall type that calculates based on the size/volume of a CUM block? Short of doing the math for how many blocks/ how much weight per unit length, and enterting in formulas?

I guess the same Q applies to stud walls... Obviously i could set up a formula saying Length/16" + 2 or whatever, to box the ends in... But what about windows? Anyone have *smart* wall types or something similar? As im typing, im thinking there may be an overlay, or an add on that does this... I know we had one for ArchiCAD, and ill do a search tonight when i get home...

I realize the entire post is pretty vague, but as im new to Revit, and im still in the learning curve, id like to see what everyone else is doing, so we can start to push it.

[Tom Dorner]

You have hit upon one of my pet peeves with a lot of models being created in Revit where people focus on making it "look" right on paper, rather than be right in the database. It is going to be a huge problem in the pursuit of getting accurate estimates and quantities down stream in the process. I run across the problem everyday in the firm I work at as the rest of the folks here in the office somehow think that the architect produces plans as the final deliverable. I argue that what we produce is the building and the plans are just one of many means available to us to describe the building with Revit.

That said, assuming a reasonably accurate building model, there are a couple of options to explore for estimating.

Probably the simplest way is to do a couple of schedules inside of Revit with some formulas to determine quantity. While being the simplest, this is the least powerful as you would have to embed within Revit the costs that are usually kept in outside databases such as RS Means. This would quickly turn into a nightmare scenario on a project of even moderate complexity.

Another option would be to perform an ODBC export out of Revit and in the database world link in that ODBC database to your cost database, develop all the required queries and reports. Again it is possible but a lot of work. Sometimes you will want to estimate something on a raw quantity, sometimes by assembly, sometimes by cost per sq. ft. and sometimes by an allowance. Or it is probable that your cost estimate is a combination of all of the above. Taking all of this into account would require a lot of effort on your part in the database world.

Another approach would be to use a commercially available estimating package such as Timberline that already has the various estimating methods and databases of cost built-in and feed Timberline with the right kind of data from the Revit model. Timberline has what they call a CAD converter module which will accept the IFC 2.0 file format. You can export out of Revit 8.1 to IFC 2x2 which is one step ahead of what Timberline will accept. So you run into a dead-end with IFC and Timberline. You might be able to ODBC export out of Revit and query properly and save to an Excel format that Timberline can read in, which is my next attempt to get Revit to play with Timberline.

I though the Revit API would hold some promise until I took the API class at AU and found out that you need to be in Revit to run any application you develop. This really limits the ability to get at the Revit data from outside of Revit. I was hoping that from the programming world you could simply "dig into" the Revit database without having Revit running. I guess a Revit ODBC database driver what what I was expecting for to establish a DSN connection direct to the .rvt file.

At AU there was a brief demo of doing an estimate from a DWF file via Excel which again doesn't hold the power needed for most commercial projects.

The long and the short of it is that while the promise of estimating and ordering materials direct from the Revit model exists, reality suggests that there are not a lot of viable options as of yet in my opinion.

Maybe this thread will spark some further discussion and ideas on the subject.

[sbrown]

editing the profile of a wall does indeed change its area. So that shouldn't be a problem. However I would not trust a model to really do take offs, Its a great idea, but not from the architects model. Thats why there is a whole new profession called Construction Modeller. This person will rebuild architects drawings or revit models, in the way the contractor wants them built. I know this sounds redundant, but I can't see how a contractor will ever trust a model he gets from an architect and I can't see an Architect guaranteeing the accuracy of his model for cost purposes. I use the quantities in revit as a check for a contractors bids, change orders, etc. It will be interesting to see how this all pans out in the long run. I think design build firms will start to rely on the model first and they will build it as they build it after the design phase is done. More contractors are now coming to owners and proposing they be hired to do the construction document instead of having the architect do them. Obviously they need an architect on board to do this, but if you think about it, contractors know how to build buildings, we as architects(not all mind you, but many) are just now learning how to build them after years of just "drafting" them.

[aaronrumple]

We have that need now. It's called Design Build. Even if we are hiring a subcontractor, we still need to know what the quantities are. Right now we have a job. Our CM department "priced" the job. We've made some design changes. They "repriced" the job. When they did the pricing, they just went to market and said what's the add/subtract for all this Mr. Sub. They didn't do the take off and pricing themselves. They should have of course, because I found several items that we didn't get deducts for (of course). If we had done acurate quanities take offs in-house, we could hammer on the sub and say "This number looks funny, where did this 1000 sq. ft. of wall go?"

[Tom Dorner]

As Scott points out, many people will not trust the architects model for estimating purposes. I agree with Scott's view that a whole new field is emerging for a construction modeler to work direct for the contractor. There are listings popping up on Monster.com for just such a position.

I'll submit though that it is only a matter of time before the inefficiency of building a virtual building model twice will be the catalyst to drive more construction to the design-build process. Then you will build the model only once with the complete end game in mind. Close to 50% of construction in some building types is already design-build and projections are for those numbers to increase dramatically in the next ten years.

http://archrecord.construction.com/resources/conteduc/archives/0511edit-1.asp

In some markets such as here in the Twin Cities of Minneapolis-St.Paul, we have a very strong acceptance of design-build with a couple of major firms heading up the effort here locally and nationally. I know these firms are busy implementing Revit as they keep showing up in greater numbers to our local Revit users group. The benefits of BIM/Revit are extremely compelling in the design-build environment.

[rod.74246]

All very good points above.

I see all these problems all the time and i can;t trust our own models for take-offs either. unfortunately it is all down to pure laziness and sloppiness on our parts so we really have no one else to blame. We have great opportunities to enhance our role in the construction process and once again we seem not to perform and open up a whole new industry based solely on the fact we aren;t doing our jobs properly.

I too am guilty of this so i shouldn't judge.

[sbrown]

I really push modeling correctly in our office, too. But currently there is little reward to an architect for modelling correctly, the only reward revit offers is getting are 2d drawings done quicker. If a contractor/owner becomes willing to pay us as the "construction modeller" some sort of sep. service. then we will be more likely to model correctly. the other problem is somethings you really can't model correctly and have to use details, filled regions, etc to have the drawing look right. this ofcourse isn't reported in quantity take offs etc.

[twiceroadsfool]

I really push modeling correctly in our office, too. But currently there is little reward to an architect for modelling correctly, the only reward revit offers is getting are 2d drawings done quicker. If a contractor/owner becomes willing to pay us as the "construction modeller" some sort of sep. service. then we will be more likely to model correctly. the other problem is somethings you really can't model correctly and have to use details, filled regions, etc to have the drawing look right. this ofcourse isn't reported in quantity take offs etc.

This is more along the lines of what im curious about getting around. Obviously the methodology and job practice issues are things that wont be addressed with the software, so ill save most of it. I will say this though... We as architects (here, anyway) have always produced our own CD's, except for specialty shop dwgs. The day i hand that over to someone else (construction modeler?) is the day i change my title to *designer*... because if all i make is a sketched out pipe dream, im not an architect anymore, lol.

Anyway, im curious more about the programs abilities for getting around the inaccuracies. For instance: A 12" CMU wall... I can make it a CMU wall, but its still really just a solid mass, with hatch on it. Granted, it will calculate linear feet, which is all i really need for a CMU wall, but you get my point. What are my options if i DONT want it to just be a solid mass? The same holds true for stud walls...

As for the Edit wall profile thing: Yes, it changes area and volume... but if you use it to change the length of the wall, the official *length* of the wall doesnt change. I guess thats a method thing tho, ill have to get people to stop doing that, lol...

[sbrown]

It should give you an error about edit elevation profile is not the way to change length or height. So one way is to allways make your walls the proper length and only edit the profile to change the shape.

[Tom Dorner]

Anyway, im curious more about the programs abilities for getting around the inaccuracies. For instance: A 12" CMU wall... I can make it a CMU wall, but its still really just a solid mass, with hatch on it. Granted, it will calculate linear feet, which is all i really need for a CMU wall, but you get my point. What are my options if i DONT want it to just be a solid mass? The same holds true for stud walls...
Your 12" CMU wall would be a wall assembly in most estimating packages. Your estimating package then does the math to arrive at a material quantity to actually order, figuring a built-in waste factor. This is why it is important to pay attention to the Wall Type Marks (no duplicates allowed) and to the CSI Assembly code (don't duplicate your 12" CMU wall to an interior stud wall without assigning the correct CSI Assembly Code). It is the combination of wall types and assembly codes that will allow the estimating packages to assign unit or assembly costs and calculate quantities.

You would not want to somehow model the individual CMU's as this would be frightful even if the PC's of tomorrow could handle it.

[twiceroadsfool]

Your 12" CMU wall would be a wall assembly in most estimating packages. Your estimating package then does the math to arrive at a material quantity to actually order, figuring a built-in waste factor. This is why it is important to pay attention to the Wall Type Marks (no duplicates allowed) and to the CSI Assembly code (don't duplicate your 12" CMU wall to an interior stud wall without assigning the correct CSI Assembly Code). It is the combination of wall types and assembly codes that will allow the estimating packages to assign unit or assembly costs and calculate quantities.

You would not want to somehow model the individual CMU's as this would be frightful even if the PC's of tomorrow could handle it.

Got ya. Sounds like a decent way to go about it... Were generating loosely used schedules for quantities and whatnot right in Revit, we arent using any estimating packages yet... But maybe thats something i should bring up to the higher ups... Perhaps thats a safer way to go.

LOL @ the computer thing. Ive never seen something kill a computer like this stuff. 3Ghz and 3GB of RAM, and its still churning over this stuff.

When we were using Catia @ my last job, because it had no architecture based standards, we were modeling EVERYTHING... Down to the rebar in the footings. We got as far as a 1 story prefab warehouse, that we used as a test run... Producing CD's for it was a disaster. The computers couldnt handle the model regenerating everytime you rotated it, lol... But damn, it WAS a cool model.