6.20.2012

BIM Metrics

The following is an email I wrote to a LinkdIn contact Tim Cole of Causeway Software. Tim had asked me about what I thought about how the Levels of Detail effect the kind of data which can be collected as Key Performance Metrics around BIM models. This is my second reply in the discussion between Tim and I and I thought others might be interested in this particular segment. The issue I'm dealing with is at the the heart of metrics in any form, and that's the data collected and for what purpose that data is intended. My contention that the kind of data some folks want to collect is a direct reflection on the authoring software and purpose of the project's original intention. So here goes. I hope you enjoy the read. And as always remember, "Collaboration is the Glue of Success."

Tim,
I'm not sure how I came up with the five levels, but they are similar to the US AIA level of detail document. I've modified that level to suit what my experience has taught me and what I've seen more advanced Owners such as General Motors, Ford, Motorolla and Intel use as their increasing tiers of service they ask for. For man of us in the US the first level of use is just assumed to be a 3d model. I know you can have a flat drawing that has data attributes on it, but it really is not a virtual building model, it is just a flat drawing with data attributes. Yes I do know you can do 3d Acad with 3d blocks and attributes, and that could be constructed as a BIM model, but it is pretty crude and rudimentary. Acad never was intended to hold a lot of data attributes, it is a drafting program. Using it in the extended form is a bit cumbersome. Then there's Catia. A premiere 3d mechanical modeling tool Frank Gehry has spent millions on to make it into a building modeling tool. And a very good one, but at a tremendous cost to own and operate.

Then we get to the built for purpose software like Revit, Archicad, Tekla and Datacad. These tools were built to handle data and intended to model space and construction from the beginning. Revit's history comes out of a mechanical modeling history but the guys quickly found that a mechanical parametric modeler just didn't work well to construct buildings, so they started all over and used a completely different mathematical base for the definition of parts. There is only one other program I know of designed like Revit since the 1980's and Tekla owns that original source code now. Tekla and Revit share some common ancestry in the algorithms they use to define elements but Revit has taken that definition and internalized the data store to be almost entirely inside the model. That, I think, is the one weakness of Revit, but eventually they will change this and make most of the data external and accessible to 3rd parties as the projects grow in size and complexity.

The takeaway here is that purpose-built software intends to create a virtual building from the outset and not be a drafting or documentation tool. That is the biggest difference in the two camps or classes of software on the market today. Here in the US most of the users don't think of any 2d documents as BIM. Maybe intelligent drawings, but certainly not a virtual building. So for us starting at Level 1 as a 3d virtual building, although it may be just a very general idea of the design and construction, or maybe no construction technology has been thought of at all. It really just conveys a design intent and not much more. You know that these models are quick and dirty and there are a lot of 3d modeling tools that can be used to get this kind of result, but it's not a BIM model if there isn't any Information component in it to help w/ completing the design. This stage is where Onuma really works well. Their space and utilization tools are really nice for building models at this level. They allow you to use something like Sketchup and then marry that spatial model to data by importing the 3d into Onuma and apply lots of data attributes to the spatial construction. There is a lot of value there to FM and Capital planning folks that don't really care about the construction detail.

The work I've been doing for the last 10 years has been focused on how to leverage the incredible resources of a virtual building process into the overall delivery methods we use. I've found that no matter what level of detail you use in a model there is a much more important story going on along side it which we leave completely out of the picture and that gap creates a lot of confusion and cost. Marrying those conversations with the model in a central storage repository so anyone connected to the project can understand why decisions were made would be a tremendous leap forward. There are so many levels where this 'conversation dialog map' could answer all kinds of questions about what is going on. Paired with a knowledge store of allied information keyed as a semantic network of relationships is what I think is needed to help BIM deliver the meta information which is really where the real value is during the building's lifetime. This goes way beyond the concept of levels of design and gets back more closely to the original questions about BIM metrics.

With a data store as described above you would know about how decisions were made, what the design assumptions and goals were, how they were intended to me met and how they were actually met in the end product. These are the metrics which are most important to Owners, not what level of detail we used in a model. The results of thoughts which turned into action, this is what building owners and investors are interested in. What was the return on investment for a design implemented over time? Did the extra money spent, really result in better service and a longer material service life?  Right now we aren't measuring this kind of performance. In fact, we aren't even actively determining if USGBC certified and rated buildings really perform over time as they were intended. There are some pilot projects going on, but a limited response for now. So do Owners really get any benefit out of a 'green' building? That to me is an important metric which few people are doing any significant work on now.

In the end, all metrics are driven by data. Data that can be reliably gathered and analyzed in a standard format (read an extension of accounting rules). Until the profession matures enough to settle on some data-driven processes, this question of metrics will be one of many words, conferences and papers and a few people who bravely go out and try something, anything to see what works. Those brave few will be the ones that set the pace for the rest of the professions years afterwards.
Andrew Abernathy
 This post is a continuing String of  articles as on the effects of Building Information Modeling and Virtual Design Construction on the Built Environment. As such there are Connections between the worlds of  design, software, economics, finance and Facilities Management.

3 comments:

  1. Metrics are already been applied to COBie as the key data deliverable in US and UK design and construction projects. Metrics measure the likely utility and completeness and accuracy of the information provided. We will also soon be assessing 'continuity' between data deliverables, as assessment of changes and deltas are also important to the client.

    ReplyDelete
  2. Nick,
    Thanks for the reply. Our conversation isn't focused so much on related data as to the metrics regarding performance/use of BIM as a contributor to the overall process. The use of COBIE as an interface would be closer to what we are referring to than the data directly related to the project housed in COBIE. I hope that helps others frame the discussion as it was intended.

    ReplyDelete
  3. Andrew,
    Something that Ralph Montague commented on, on CITA BIM LinkedIn group a while back. Very nice diagrams in Laura Lee's report for South Australia state. The download link is a bit slow, but it works.

    Lee, L. (February 2011). An Integrated Design Strategy for South Australia: Building the Future. Adelaide: Department of the Premier and Cabinet, State of South Australia.

    Regards,
    Brian O' Hanlon

    Ralph wrote (3 months ago),
    Professor Laura Lee, author of “An Integrated Design Strategy for South Australia - Building the Future”, addressed the RIAI conference in RDS Dublin yesterday.
    See a copy of the report here:
    http://www.thinkers.sa.gov.au/lib/pdf/LLreport.pdf

    ReplyDelete