Wednesday, October 15, 2008

BIM Content - Part 2

So today I’ll try to wrap up my thoughts about “BIM Content”, although it won’t be the last time you’ll hear about it :) What am I trying to get to? I think it is quite clear that to create suitable content, you need to know what the end use is. Is it a piece that will be suitable for all aspects of a project or a series of interchangeable parts that are replaced at different stages throughout the life of a building project? If the latter is the case, then there's a lot of coordination that needs to happen to safeguard against elements moving in relation to each other when these parts are replaced, or to protect the integrity of the data from being lost or tainted. How will this coordination be handled when we use third party content?

Yes, I'm one of the skeptics when it comes to using manufacturer supplied content; at least for now I am as I see too many variables. We have a hard time coordinating efforts within our firm and we're the ones building content to meet our specific needs! Most of the content that we get from most web sources ends up being unsatisfactory in terms of function, quality, level of detail, etc, or it lacks the 2D representation that we’re after. And how can we trust that this content represents the latest version of the manufactured component?

Navigating through content in Revit’s Project Browser leaves a lot to be desired. It is very cumbersome to find what you need in the Type Selector especially when using the Component tool to place an object. Currently users are forced to come up with naming conventions in order to group content logically or abandon the use of the tool and use the drag & drop capabilities directly from the Project Browser. The interface can help us a lot with sorting out the categories for example, instead of showing us big, alphabetized lists. After all, it is a database! The same applies to scrolling through the expanded tree of categories of families and types (don’t you hate it when colleagues expand every single node?!). On small projects this might not be too bad, but when the size and complexity of projects increases, the volume of required bits and pieces grows exponentially and has adverse effects on the efficiency and productivity of the team. Currently I personally focus more on minimizing the amount of unique families needed in a project and try to manage their configurations and dimensions through Types and instance parameters. This can lead you to a point where families get quite complex and more difficult to change later. Regardless of how snazzy content navigation gets, I think there is a lot of value to be gained from being efficient in how you build content, at least while we’re still in (1) and (2).

There’s going to be a big leap that the industry needs to take to get to (3). I’m only focusing on the content aspect here and staying away from the legalese, etc. Assuming the use of a digital model moves from phase to phase and the information in it is used, refined and enhanced until we end up with a real building and an identical digital version, it is clear to me that the content in the beginning stages would differ from that in the operation stage. Not to mention that during this stage, some components might be replaced with different ones. How can we safely replace content as the digital model moves across this “conveyor belt” (time)? It sounds to me that unless we have some form of “syndicated” content, then we cannot realistically get to (3). Should the different manufacturers be the ones that supply it? Who sets rules/guidelines as to how this is built? What happens with building assemblies that are not put together in a shop, but built on the jobsite (think of drywall walls and ceilings for example)? The implications of variance and lack of accuracy change with each different object in a building project. For example a discrepancy of an inch in a wall mounted object will not have the same effects as a cooling tower that is larger and heavier than the one used as design basis to size the support framing. For BIM to work properly, we need access to content that we can trust. Dimensionally speaking, trust has to reach its maximum as we start construction.

Our current reality is different but I think we cannot ignore these questions and we need to work towards a successful framework. If we move completely towards digital models, how will detailing decisions that we currently illustrate in 2D be communicated? If we intend to just work in 3D and print to paper until I reach retirement, then we should be fine I guess! With our current process, there is a certain “sketchiness” built in: the builder uses a measuring tape to transfer measurements to the slab for build a wall (yes, some are using digital instruments too, but not in all stages of construction) and most elements are built on the jobsite after measuring partially built assemblies, etc. But if the promise of BIM is to be fulfilled and we get to a point where we “blindly” trust the digital model’s accuracy and have more complete components delivered to the jobsite for final installation, then that content needs to be trustworthy and adaptable/exchangeable for downstream use without the risk of data loss/infection. We’re also assuming that the A&E industry is willing to participate in post-construction downstream use (Operations). Is that a realistic assumption or a gross one? If we’re not willing to as an industry, then why would we worry about how the content is replaced/adapted for future use? Should BIM stop at (2) for the construction industry and pick up from there for Owners, Operators and others interested in offering building management services?

I have a lot more questions than answers on this topic and am very interested to see what path we’ll take in the immediate future. In the meantime we obviously need to focus on efficiency and productivity gains from our content (ex: 3D Modeling needs to yield the exact 2D representation we require for print as we cannot afford to spend too much time “cleaning up” to achieve both). We struggle with wanting to be accurate in our modeling, but if something doesn’t represent well in 2D, we stop short and end up with incomplete and inaccurate 3D information due to a shift in focus towards traditional 2D illustration. We cannot afford to end up with untrustworthy 3D models and “ok” 2D print hybrids.



Post a Comment