It was mayhem in the hours leading up to Smartgeometry 2013’s exhibition on April 19. After four straight days of working from 9 a.m. to midnight, the 100 participants in the 10 clusters, or expertise groups, were running on adrenaline to put their notion of the future tools in architecture and planning on display. Laser-cut remnants, tangled power cables, and mortar dust were cleared out of the Wates House at the Bartlett, UCL’s faculty of the built environment, as participants set up a mix of presentation media—projectors, models, and even robots—to communicate the ideas and issues that they tackled for this year’s challenge, “Constructing for Uncertainty.”
As part of the press tour organized by Smartgeometry sponsor Bentley Systems, ARCHITECT explored several of the teams’ exhibits. Within the past 24 hours of the last active workshop day, participants had transformed the screens of abstract code, digital models, and computing engines into planning and design solutions, interactive design tools, and even physical objects. Even more exciting is the notion that the capabilities and technologies created in the past four days represent a taste of what the future of design, planning, and construction may become in the near future.
Smartgeometry participants and attendees mingle on the lower level of the Wates House at the Bartlett, UCL Faculty of the Built Environment in London. As the fabrication and 3D printing studio, the lower level of the facility saw a frenzy of actvity of a different kind in the days leading up to the April 19 exhibition.
A test façade panel from the Thermal Reticulations cluster, which compared the thermal performance of different façade geometries and materials against the performance predicted through digital modeling. The team, which hailed from the Royal Melbourne Institute of Technology, wanted to investigate the uncertainty of where the physical and digital results would intersect.
In Projections of Reality, the cluster from ETH Zurich augmented data obtained through 3D capture and scans of physical objects with digital algorithms that predicted properties such as solar incidence, walking distance, and viewing angles. The data was then projected back onto the physical models. Participants could move the foam blocks or use their hands as solar trackers to interact with both the digital and physical models in real time.
The Adaptive Structural Skins cluster cited origami as one of its inspirations for building a 2-meter-square, retractable and continuous skin, which the team first modeled digitally.
The Robotic Foaming cluster filled its exhibit space with structures made with robotic arms that followed actions scripted by participants.
The hand-mixed composite material, made primarily with polyurethane, was a significant uncertainty in determining the structural success of the resulting forms.
The Computer Vision & Freeform Construction cluster used digital guides to build two masonry vaults, one without the aid of an experienced craftsman. By frequently checking the position of the form against fixed markers on the wall, the group compared the built forms with the ideal modeled forms.
The Digital Intuition & Prediction participants used generic algorithms to conduct a series of building investigations that looked into pinpointing the optimal shape of a performance hall, a structurally efficient form, and building geometries that could maintain prescribed program adjacencies.
Through coupling probability theory with parametric architectural and urban design, the Probabilistic Architecture Design (PAD) group sought to identify the optimal structure heights and forms for a given goal, such as a high amount of sun exposure with the least amount of solar temperature gain. The team used Bentley's GenerativeComponents software to model the different possible solutions.
The Volatile Territories cluster used algorithms to accelerate the spatial redistribution of programs due to changing contexts. The team translated three-dimensional Cartesian coordinates into RGB colors to represent the outputs visually.
