Friday, August 20, 2010


Blog entry by: Dominik Holzer

If one investigates the information-flow of traditional architect - engineer collaboration over the last century one will most likely encounter a process where engineers are asked to join a design project at a certain stage to help realize the architect’s ideas. The shortfall of such a work-method is the relative disconnectedness between performance feedback and design morphology.  Those architects and engineers who aim for more integrated design-practice are in search for ways that allow them to design more concurrently and to exchange information bi-directionally.  Such a process requires a high level of skill, trust and confidence in those involved. It forces architects to let go of their role as sole-authors to admit outside involvement from consultants. It requires engineers to become more proactively engaged in the design process and it prompts them to consider the bigger picture of how their input affects the overall design, both technically and aesthetically.

The engineering firm Arup has been fortunate to be involved in a series of exciting projects where the design team has searched for a more integrated approach. Some recent examples of this collaborative effort include the Watercube project in Beijing, the Marina Bay Pedestrian Bridge in Singapore and the Melbourne Rectangular Stadium in Australia. 
Figures 1 -3: Watercube, Marina Bay & Melbourne Rectangular Stadium project examples
Computationally assisted building analysis and simulation has revolutionized the work-methods of engineers and it is impacting drastically on their capability to respond to design alterations with according performance feedback. Depending on the type of analysis required, some performance feedback can even be provided in real-time, thereby closing the loop between morphological experimentation and engineering-performance checks. New tools offer new possibilities for collaboration, but they also force architects and engineers to reconsider the planning and design methods they apply individually and in teams. Those design partners who wish to collaborate closely need to put stronger emphasis on defining the over-arching design drivers and the performance-parameters associated to them at the outset of a project. More than that, a new method of engagement is required to evaluate and trade-off multiple design options in a short matter of time. Optioneering is such a method.

Experience gathered by Steve Downing and me during our collaboration at Arup has shown that optioneering is primarily a method for collaboration which does not necessarily depend on any specific software. Optioneering logic implies a form of understanding between collaborating parties who discuss the criteria space for a design of a specific problem at the outset of their collaboration. In doing so, optioneering provides the design team with a common ground to discuss the design and understand the effect of changes on each other’s discipline. Architects and engineers can then address and trade-off the criteria through a controlled set of alterations that yield informed choices for decision-making. 

Optioneering processes are intrinsically tied to human decision making. Although the search for possible solutions through optioneering benefits from automated processes for design evaluation, the ultimate goal is to provide the design team with an array of possible design options for collaborative prioritisation, selection, and decision-making.   

Figures 4 & 5:  Diagrammatical  representation of an Optioneering workflow, Potential optioneering Network

In order to support optioneering, two techniques have proven to be particularly useful in our research: Parametric Modeling and Multi-criteria Decision Analysis (MCDA).
Parametric modelling tools help designers to produce a quick turnaround of design options by allowing users to generate multiple design alternatives (Geometry Cases) to keep a design in a flexible, yet controlled state. MCDA tools for engineering design optimisation solve complex trade-studies using optimization algorithms and Design of Experiments (DoE). These tools are rather generic as they support many kinds of MCDA and they do not necessarily focus on the building industry.

Dominik Holzer (AEC Connect) and Steve Downing (Arup) published an extensive account about the Optioneering process at Architectural Design, Volume 80, Issue 4, pp. 60 - 63.AR, John Wiley & Sons, Ltd.

No comments:

Post a Comment