Tuesday, November 22, 2011

Content Critique

Content is King! The right content, good content, smart content...

How the content is to be judged is subjective, evaluated against personal criteria, as well as objective. It doesn't take a lot of effort to find something to complain about when you use content created by others. It isn't even all that hard to return to our own content a little later and have a different opinion about how well we made it. Ever evolving, improving...hopefully.

I had the occasion recently to poke and prod some families shared via Autodesk Seek. The first example is from Belden, one of their equipment management products, this image is a 3D view at 1/8" = 1'-0" scale.

Another example is from Siemon, one of their punch down patch panels, also in a 3D view at 1/8" = 1'-0" scale. This is a bit blobby eh?

What you see in these images are what you see in the project environment, assuming a particular scale as mentioned before. There is a heck of a lot of detail in each of the families. The rack weighs in at 796 KB and the patch panel weighs in at 2.056 MB.

Taking a closer look at the punch down patch panel we find this, I had to use thin lines to see the detail.

The back side has the punch down blocks modelled too as well as sheet metal bends and kerfs. There are even small parts on the inside of the panel which nobody will ever be able to see in a Revit view.

Another family I had a closer look at is one of the Siemon Wire Management racks which weighs in at 2.9 MB.

It is nice looking but there are elements that have been modelled that nobody can see in a Revit project. Cutting a section through the cabinet will not yield the extra modeling effort that was put in. A simple rectangular shape would yield sufficient results in nearly all views except for a close up photoreal rendering perhaps.

These last two families don't take advantage of Revit's Detail Levels or Visibility options to manage the complexity or detail they contain. What's more troubling about these is that they are posted on Seek, suggesting that they are ideal or represent content that others should emulate. The wire management rack is the most "over the top".

I've hidden a hundred or more elements (solids/voids) to show this stuff.

Nearly everything you can see in that image is hidden behind something (except for the door handle) which means in a Revit project that nobody can ever see it because the category (Electrical Fixtures) doesn't have a cut representation. In a few cases I've run across data related equipment that have been assigned to the category Specialty Equipment, which doesn't even show up in Revit MEP views ordinarily (not included among the MEP categories listed in V/G, without showing all categories).

Where am I headed with this critique? It's still the wild west folks. Just downloading content from Seek is no guarantee that you won't have to spend some time tweaking or in the case of the wire management rack or punch down patch panel, abandoning them to create something simpler.


Bruce Madsen said...

In my experience, this is the sort of modeling is all too common from manufacturers.

The manufacturer’s goal is to represent their version, and their “content creator” will do it for them.

Siemon is not the only example of this misguided application of resources.

In my opinion, it is a common problem for "content consolidators" with a business model based on “services for manufacturers”. (Content consolidators are Arcat, smartBIM, and the like.)

It is sad to see that SEEK has abandoned their own guidelines and allowed this content into their database.

Parley said...

Come check out our class at AU. We'll be discussing this very topic... from the stand point of a SEEK Content Service Provider. There’s no doubt, there is a lot of garbage legacy content in there. SEEK has recently made HUGE inroads as far as cleaning this up, with more to come very shortly. In the above Belden example, (we made it) the file size is ONLY due to the model text which is only visible in 3D. This was a feature that was requested by actual RCDD designers. We feel that BIMAdvent’s approach is unique in that our content is completed with the designers in mind.
We'll also be discussing the performance ramifications of file size etc.. It'll be insightful. Great post.

Steve said...
This comment has been removed by the author.
Steve said...

In fairness, since I wrote that none of my examples, the Belden rack does use detail level and visibility options. The 3D text shows up in front/back elevation and 3D views but only when using a detail level of Fine.

Removing the 3D text and logo reduced the file size to 468 KB. However, I've always written and acknowledged that file size isn't the only measure of how well a family serves us.

tucker_arch said...

It is a Catch22 for manufactures as they want their product represented accuratly, but I think most really want their content used in models also. I have found several examples of content having too much modeled.

A recent one was a toilet partition family, the manufacture’s family was around 2.5mb while the one I created was around 750kb. Both could represent a multitude of stall configurations, but the one I create also could represent Ceiling Hung, Floor Mounted, and Floor Mounted/Overhead Braced in that one family. The manufacture required 3 separate families for the same configurations.

While ‘bloated’ content is a common issue among manufactures, I think the bigger issue is the ability to use it all together. Some manufactures choose to use the Seek Guidelines (shared parameters) while some others are only utilizing family parameters. This makes it really hard for our company to harness the “I” in BIM.

I also appreciate when content is developed with supporting content/settings as it helps the user understand how to harness the information in the content (hopefully there is some). This may be schedules, filters, or just user guides.

Steve Germano said...

This is a good point Tucker, if the manufacturer / content creator does not provide a schedule its up to the user to parse through the meta data to schedule that family. This takes too much time IMO and is a primary reason at BIMAdvent we supply schedules with every piece of content we create. We try to make your life easier in any way we can!

danielgijsbers said...

You see this sort of modeling in all kind of applications.

I once had an Autocad 3D file that was torture for any machine. I found out that the key in the door even had a profile in it...

Rahul Shah said...

I agree with tucker_arch "It is a Catch22 for manufactures as they want their product represented accuratly"

It is a real challenge for manufacturers while embracing BIM. I think the issue is their BIM/LOD is different to designr's BIM/LOD.

Please check the following post for more detailed comments on this. I wanted to explain with some images so I thought I better write on my blog.


The BIM Man said...

I too agree with "tucker_arch", it is a catch 22 for manufacturers but the smart ones should consult the industry at large for "honest" end user feedback on how best their content is to be used by the end user.

I have done some consultancy work for a content provider in the past and coming from working within the industry and a fulltime Revit user of the intended content, it allowed me to provide prospective manufacturers with the knowledge of not only if their products should be “Revitized” but better still how best to build it and how it would be used within the industry it was meant for.

I get the feeling from some of the content I’ve seen on the web to date, there are content providers out there just “cashing in” on the BIM wave and providing content that

1. Isn’t even needed.
2. So highly detailed it’s unusable especially on large scale projects.
3. or worse still just CAD files linked into a Revit file giving the appearance of being a parametric family when in fact it’s nothing more than a bunch of lines and surfaces.

I hope some useful guidelines will emerge sooner rather than later so manufacturers BIM content doesn’t go down the same path as a lot of the legacy CAD content out there and being “all over the shop” in terms of standards, levels of detail and accuracy that we all end up building the content again over and over again...

KArch said...

I also agree with comments by "the BIM Man" and "tucker_Arch". I found this post while trying to find a solution to some of my own custom built family models. I've been using Revit for the past three years and only recently have been building custom families. I tried as long as possible to use the generic AutoDesk provide families. As an architect I find it's hit or miss when seeking quality built models by manufacturers. I want and need the ability to change the dimensions, and add/remove accessory items in a single model that is compact. These features are very rare in manufacturer built families. I find the "generic" Revit out-of-box families very limiting, and the manufacturer models too often overdone. There seems to be no middle-ground, so I've started looking for well-built third party solutions or learning to build my own families as needed.

AutoDesk needs to quit giving us new yearly releases without doing a major overhaul, and improvement to their existing generic families most of which have been around for a longtime with no changes. As I've learned to build my own families I've also learned how basic, and useless beyond schematics the AutoDesk generic family models really are.

I've watched over the past 2 1/2 years "Revit-Content.com" a third-party developer as they've built and improved their generic families. I will probably buy their window module first and test it on some projects next year. The software appears to do what a designer needs, build a complete model that can be tweaked with many options to each family which appears to work in concert with there other modules. What I mean is every door, window, cabinet family, etc appears to have consistent command prompts, terminology, and functionality. As an architect I shouldn't be consumed with spending my time building families for "standard" building components. AutoDesk, doesn't understand this, period. It shows in their twenty-five year development of AutoCAD Architecture, which suffered for many of the same problems. I remember waiting for years for improvement to the block libraries and we finally gave up and made our own.

The reality is many manufacturers are clueless to developing Revit family models because they often have their own in-house CAD/CAM software and they don't understand the value of developing Revit models for designers to use.

A good family model should give the designer the capability to select desired family type, insert it, size it,then select any associated parts and components needed in the design while remaining compact, and easy to manipulate.