In his inaugural lecture at the University of Leipzig in 1893, art historian August Schmarsow redefined architecture as a spatial art. This represented a major break in (‘break in’ implies that architectural thought ceased for a period of time. Needs rephrasing. Do you mean a ‘break away from traditional architectural though’?)  architectural thought, as since the Renaissance, architecture had almost solely been discussed in terms of style or construction. When computation was introduced to architecture some 60 years later, its potential to provide new representations of processes and forms of occupation were quickly understood, as presented by Christopher Alexander, as a theoretical framework to systemic design processes,2 or Archigram’s speculations on new spatial types.3 Alas, its more obvious powers to crunch big data, optimise explicit parametric problems and accelerate existing procedures have sidelined most developments that attempt to deal with new representations of space and use. Major contractors such as developers or public planning bodies reinforce this trend by demanding building information modelling (BIM) as standard (reducing the choice of tools to generally two software providers), while parametricism is nowadays perceived as state-of-the-art computation. Both serve the purpose of automating traditional professional deliverables.

As a medium for automation and optimisation, contemporary design computation is more aligned with Modernist thinking where the machine analogy dictated frictionless operations of closed spatial objects either as buildings or city masterplans. Unplanned human interference was designed out, leading to cognitive dissonance and, ultimately, dysfunctional use. The other current role of architectural computing, reminiscent of mid-19th-century eclecticism, is associated with decorative additions wrapping conventional spatial organisation. Best-known examples, mainly from academia, extol pavilions displaying no spatial complexity whatsoever.


This is an odd anachronism as building and urban planning professions have changed their approach in favour of user-centric spaces over the last 15 years – at the same time as computation has become mainstream. Best Practice design guidance such as the Council for Architecture and the Built Environment’s ByDesign4 or WorkPlace Matters by the General Services Administration,5 recommend qualitative design objectives that appear difficult to quantify. New transport stations such as those for the London Underground Jubilee Line are designed to enhance perceptive performance for legibility Figure A1, offices are designed to increase social interfaces (for example, all Google offices) Figure A2, and cities for sustainable use (Masdar City). This new design approach builds on the awareness that better spaces must be based on users and their relation to spatial organisation, on social interaction and human cognition as drivers to achieve better performances.

Such qualitative user-centric spatial issues have been the domain of a less visible architectural computation agenda that is slowly surfacing as spatial performances and user behaviours become increasingly important. Buildings such as shopping centres, transport stations, international workplaces, hospitals or schools are increasingly complex and planned around two dynamic drivers: operational schedules and the behaviour of their (many) occupants. Forecasting and controlling these drivers provides effectiveness and robustness, but also attractiveness in use, meaning that the design process has to accommodate dynamic patterns by understanding user behaviours that also affect operations. The drivers are progressively weaving buildings or neighbourhoods into larger more complex systems of spaces and eroding the boundary between places. So far, user-centric simulations are generally post design and driven by the need for evidence reporting, such as pedestrian simulations for fire egress. But the abovementioned, less visible computational agenda has asked exactly those questions of the correlation between spatial dynamics and form. This issue of AD discusses different strands of this alternative development, the (architectural) theories that have informed it, and current state-of-the-art implementations.

Read full paper here –