It's been more than two years now, since Autodesk started to invest in research on Generative Design. Some of the results have been shown already in the video below.
When we talk about Generative Design, the subjects are very often related to manufacturing (design of heat exchanger, bicycle frames, …).
At Autodesk, we want to show that this technology can be used in the world of AEC as well.
After having bought three levels in the "MaRS Discovery District" building in Toronto, Autodesk has used the technology of Generative Design to perform a space planning (like office space, meeting rooms, utilities, …) in a pragmatic way. The team of "The Living" (an architectural research team within Autodesk) has carried out studies concerning the initial conditions of the levels, to define the best way to meet conceptual objectives in the new building area.
Then, after having interviewed the future users of the work spaces, several wishes and demands have been collected and been analyzed with diagrams like the one displayed below:
The team of "The Living" created a series of formulas for each goal, to get several space planning solutions. Therefore they created mathematical models based on equations to characterize the objectives.
Various requirements have been assigned to the development zone, such as the proximity of teams and the distribution of equipment, such as kitchens and meeting rooms. And finally also the placement of the circulation areas that connect the several rooms.
With the help of a general algorithm, they've harmonized all the equations to find solutions that meet the objectives. This algorithm is composed of these 4 big steps:
- Generation of data (geometries, position of utilities, spaces, …)
- Evaluation of each objective
- Production of all solutions meeting the several requirements (brute force technique)
- Analysis and choice of the best solution(s)
In the final step, the solutions are presented in a Kiviat diagram, better known as a "spider diagram", to learn more about the different components and their obtained scores within the optimization.
Important here is that the optimal solution in the "spider diagram" is the one with maximal cover (indicated in blue).
Though this is a very exciting project, we still want to remind you that this is about a concept, that is not yet finalized, but which creates a lot of opportunities. For example, take the case of a residential building in which one wants to define a series of constraints, such as:
- Orientation (towards sun, environment, neighbors, …)
- Buildable Area
- Proximity of one space to another (ex. Living room close to the kitchen, but far away from the bed rooms)
- Cupboard area (minimize them)
- Maximize the exterior view
- Daylight optimization (to minimize the number of artificial lights)
- HVAC regulations
At the end of the calculation, Generative Design, proposes several solutions, respecting the requirements (criteria). This is very beneficial for an architect to define the final and optimized building model.
Originally posted by Olivier Bayle (Autodesk) on the Village BIM blog