image by archi-lab
image by archi-lab

Semi-Automatic | Computational Design | Konrad K Sobon | 2011-2013

Semi-automatic is an investigation into automation processes in architectural design. It’s my personal exploration into the world of computational design in architecture. It has long ago occurred to me that the advanced technologies used in design currently offer not only ability to speed up the process, but also opportunity to look at architectural design from a whole new perspective. In my opinion architecture and architects are no longer the organizers of spatial conditions; architecture no longer exists in the static world of spatial relationships. Architect’s role shifts from that of an organizer of static spaces into that of a designer of complex space systems. Architectural space no longer requires having a predetermined function and location. Paratonic systems have replaced the traditional space with their ability to adjust to constantly shifting needs of a user. Architecture capable of gathering, processing and acting upon data streams is what I have attempted to investigate in “semi-automatic”. There are also other, more conventional approaches and uses of computational design that I have investigated, that are being presented below.

Parametric Form Finding – Street Shading:

This project was a response to a street shading challenge. The idea was to develop parametrically optimized tensile shading. I have developed a Grasshopper definition that used Kangaroo live-physics engine to run a form finding simulation. In order to truly optimize the forms as well as to generate variation within the family I have used Galapagos and Geco to run a evolutionary simulation of which goal was to maximize the Incident Solar Radiation and in effect maximize shading at the street level. This entire process was automated using Grasshopper and multiple plug-ins as well as Ecotect for solar analysis. Formal idea for the shading devices was inspired by Sequential Photography of a street dancer. It expresses the true nature of changing dynamic that can be found in nature and architecture.

Program Blocking Informed Facade Design:

The idea was to create a parametric process to bridge program stacking diagram and facade design strategies. I was able to create a Grasshopper definition that drives certain facade parameters based on their proximity to a block of program. Program is defined as boxes that each have an “opacity” parameter assigned to it, dependent on its function. For example a lobby has 100% opacity as one wants that to get the most natural light. Other programs like Storage would get 0% opacity to keep them closed. This tool allows me to quickly shuffle program around while making sure that I don’t have to remodel the facade each and every time. It assigns values to program based on Rhino layers that they are on. Geometry modeled as facade was inspired by a shark’s gills. Objective of this exercise was to streamline some of the most time consuming design workflows. This helps quickly visualize the facade even when building is continuously undergoing planning changes.

Paratonic Space Design:

Paratonic space design – the idea is to create a space capable of gathering and acting upon its user input data. This pavilion is not only capable of direct response to physical input from embedded sensors, but also from social network sites such as Twitter or Facebook. Set up to monitor certain Twitter feed this pavilion can alter its physical apperance in direct response to a Twitter post. This experiment challanges the idea that design is inherently rooted in space and time. Allowing for communication /control via mobile networks opens up a possibility of design existing in multiple places and and across many time zones at the same time. It loses its spatial and social context. Created in Grasshopper, operated by Arduino, controlled by anyone via Twitter/Facebook; it’s a new kind of space. This design was an investigation into the changing role of an architect into more of a  system designer rather than spatial organizer.

Screen Wall:

This project was a proposal for a Medical Office Building in suburban New York. My involvement was limited to providing Computational Design assistance with the design of interior screen wall (brick wall on the images). A custom Grasshopper definition was created in order to generate a series of openings in the wall. Pattern was controlled by multiple attractor points and projected onto a surface. Idea was to create a screen that would transmit light from the well-lit and south facing lobby into Cancer Center located on the other side of the wall. Brick was used as the material of choice as to weave the old building to the south with the new MOB Tower.


Leave a Comment