We describe our work on providing support for design decision making in generative design systems producing large quantities of data, motivated by the continuing challenge of making sense of large design and simulation result datasets. Our approach provides methods and tools for multivariate interactive data visualization of the generated designs and simulation results, enabling designers to focus not only on high-performing results but also examine suboptimal designs’ attributes and outcomes, thus discovering relationships giving greater insight to design performance and facilitating guidance of further design generation. We illustrate this by an example exploring building massing and envelope design (fenestration arrangement and external shading) with simulations of daylighting and heat gain. We conclude that the visualization techniques investigated can help designers better comprehend inter-relationships between variable parameters, constraints and outcomes, with consequent benefits of: finding good design outcomes; verifying that simulation results are reliable and; understanding characteristics of the fitness landscape.@en

Issues of control and command in architecture are considered here via reflections on recent and current research projects concerning digital technologies. The projects’ topics cover a range of scales and approaches, from the planning and design of urban ensembles to the detailing of panels for constructing freeform building envelopes. Additional topics on this spectrum include methods to support open-ended design explorations, goal-driven optimisations, participatory design and the internet-of-things. In each of these the possibilities and methods for controlling the design process and the resulting artifacts and systems are addressed in different ways, which consequently influence the roles of architects in different ways. Overall we see that while digital technologies do indeed enhance architects’ control in some cases, some applications require sharing of control with others, while still others may result in loss of control either to other parties due to transferability of skills, for example or altogether due to complexity and feedbacks. Awareness of these different possibilities may aid better use of the technologies.@en

In urban planning exploration and analysis assist the generation, measurement, interpretation and management of the modelled urban environments. This frequently involves categorisation of model elements and identification of element types. Such designation of elements can be achieved through attribution (e.g. ‘tagging’ or ‘layering’) or direct selection by model users. However, for large, complex models the number and arrangement of elements makes these approaches impractical in terms of time/effort and accuracy. This is particularly true of models which include substantial numbers of elements representing existing urban fabric, rather than only newly generated elements (which might be automatically attributed during the generation process). We present methods for identification and categorisation of model elements in models of existing and proposed urban agglomerations. We also suggest how these methods can enable exploration of models, discovery of identities and relationships not otherwise obvious, and acquisition of insights to the models’ structure and contents which are not captured, and may even be obscured, by manual selection or automated pre-attribution.@en

The paper discusses geometric-content-based search and classification techniques proposed in order to support collaborative, multi-disciplinary work on building design projects which utilize three-dimensional digital models (e.g. BIM, CAD, CAE, etc.) The rationale for employing such techniques – as an alternative to conventional assigned-attribute-based ones – is presented, based on observed difficulties experienced by users of digital models prepared by parties other than themselves. Also introduced are proposed methods drawing on the field of visual analytics for displaying the results of search and classification operations by augmenting the conventional system of views and tabular information in order to enable users to obtain greater insight into the contents of complex models. Included is a description of preliminary tests undertaken to assess the validity of some of these proposals, and results showing that model users are able to apply the proposed methods for retrieval and organization of information in models, thus overcoming some language-based and other common obstacles typically encountered with exchanged models in collaborative design processes.@en