The future of CAAD? Or at least my personal take on it

CAAD is not dead. But the word "Computer Aided Architectural Design" has been used for so many things, that it is hard to not have some kind of preoccupation with the meaning of it. While BIM is currently one of the most interesting approaches, it still is only a subpart of CAAD. Just like parametric design or semantic web can fit into it. But what I really wanted to share with you is a wish list for the future of CAAD.

  • Let us choose any modeling approach and mix them in the amount we want
    • Direct modeling, such as in SketchUp
    • Parametric Modeling, with support for coding and graphical creation/editing, such as in Grasshopper or (to some extent) in Revit
    • Start from measurements, e.g. pointclouds, surveyors data, geo-information, photographs...
    • 2D drafting if you want
  • Let us work on ANY design phase and on ANY scale level in the order we want
    • in short (well, I did a PhD on that): we want to be able to elaborate rough models, but at the same time to be able to look at technical details and go back to the Master plan when we want. Ensure all information is tied and synchronized on all levels.
    • proper approach to define design variants, design phases (programming-sketch-preliminary-construction-management-end-of-life), building phases (existing-demolish-new phases)
  • Everything fits into a BIM workflow
    • objects have semantical meaning
      • a wall is not a scaled Cube with a Boolean subtracted void
      • space are real design objects, not just some attribute or fill. You can model a whole design just with spaces and define enclosing physical elements later, if you want
      • don't limit is to only vertical walls and only horizontal floors. If I want freeform geometry, still ensure that I can embed semantical meaning to it
    • design intent is captured and kept, e.g. relationships between elements are embedded
    • all partners can collaborate on the project model, e.g. with BIM servers
    • No need to ever create a 2D drawing by hand anymore
    • all calculations, simulations, evaluations can be done directly from the information in the building model
  • work directly in Open formats, such as IFC
    • make a DWG or PDF only as output for people needing that
    • reference other formats as underlay/overlay - always up-to-date

Beware, I'm not saying that all of this should be part of a single application or run by a single developer. Let competition exist, give us choice but make everything fit into our workflow. E.g. let people chose a modeling system and combine it with another rendering system and tie it to another simulation system and still collaborate with others who made different choices.

Utopic? Probably. But I'm allowed to dream.