Sunday, May 22, 2011

“Predictive modelling is the process .... which a model is created or chosen to try to best predict the probability of an outcome.”
states Wikipedia.
Finding this definition was a sad discovery for me – “Predictive Modelling” is a perfect description for a concept within BIM, yet not well-explored but with significant potential.

IT, the insurance industry, archaeology and marketing have all put their pretty-paws on this terminology already but cannot claim exclusivity.

Predictive Modelling within AEC’s BIM will be a set-of-tools developed by manufacturers and suppliers (M&S) closely associated with AEC.
These tools will work like Predictive Texting;
Within a digital modelling environment, as a user created a digital building, the intelligent PM automatically and in real time compared all of the possible forms, shapes, sizes and combinations against a built-in dictionary of elements, and determined which would possibly suit the location/conditions best.
If it guessed incorrectly, the user would get the option to scroll through other possible types, originators without disruption to thinking processes.

An idea too far out?
There is of course the danger that it gets nipped-in-the-bud by those that like to promote the design process as totally unpredictable and random.
I believe PM in BIM has legs and when M&S of building materials and products decide to finally take an active role in how building designs are developed, documented and built, it will definitely fly (high).


  1. Predictive modeling could take on a number of forms. One would be a rules-based form that allows a team to select from a number of baseline design directives and allow the software to present model alternatives from that list of 'acceptable' alternatives. Where code restrictions come into play for things like travel distances and area separation, a library of rules would assist in the design alternatives. Solibri is already doing some of this with their model checking software. Integrate that capability into BIM and Predictive modeling becomes a reality in a basic form.
    Interdisciplinary teams could use this rules-based method to determine clearance between various classes of materials as one example. If we tied energy modeling to the design we would get real-time, or batch calculation responses for each design decision for materials used, systems chosen or building orientation. Add a lifecycle cost parameter for performance of various classes of materials or equipment and the TCO of a building gets refined as the virtual environment is more clearly defined.
    Now you are going to ask, who is going to do all this as one solution? In my mind and others in the SW business, it won't be one company, but more likely an open-source solution working through middleware 'glue' that allows API interfaced software to 'hook' into other software and pass data back and forth.

    Is this vision possible? Yes. Is it probable? Maybe. Who will drive the solution in a direction like this? Owners.

  2. Hello,
    We facilitate the provision of independent analysis to support expert testimony, regulatory or legislative engagements. Frequently, this work includes economic, financial and statistical studies of varying data analysis, technical and

  3. This comment has been removed by the author.

  4. It’s great to come across a blog every once in a while that isn’t the same out of date rehashed material. Fantastic read.Surya Informatics