Free Studio Seating
Studio Seating Re-organization Optimization Project
Siye Huang, Siyu Xiao, Xinan Tan, Yuening Jiang, Yutong Deng | Spring 2022 | Generative Design | GSAPP
Introduction
The current condition of classroom seating arrangement is in a linear organization, meaning an central axis of major circulation, with branches to each of the studios. This organization ignores the privacy between students and the collective relationship within each studio. Also, it minimizes the possibility of interaction, communication and creativity. It is difficult for students to discuss while giving individual spaces with only one long desk.
Therefore, our project intends to reorganize the seating for the studios, in order to enhance the possibility of communication and interaction. We provide three different types of desk: single, double and triple, for different group structures (individual work, 2 people or 3 people group work) as basis for our rearrangement.
However, it is still difficult to quantify the term “privacy” and collectiveness”, and the traditional way of putting desk manually might be subjective and less effective. Therefore we introduced Grasshopper and Discover to optimize the seating organization with certain parameters: distance and overlapping area. The larger the overlapping area is, the more collectiveness the area could get, or vice versa, the smaller the overlapping area is, the more privacy the area could have. With these two different opposite trends that control the whole optimization process, we believe that there will be a balance in between for the best arrangement for the students future use.
Methodolgies
A series of points are set in the studio classroom plan, for generating the studio group area, the desks and their movement range. Also using sliders as input to control the ratio of three different types of desk, so that we could have an optimized number for the desks. After these process, we start our optimization process by minimizing the overlapping area between desk movement ranges (for privacy) and by maximizing the overlapping area between desk movement range and the studio group area (for collectiveness) to reach a balanced state.
Our whole process is separated into 3 steps as introduced below:
Step 1 Optimizing the area and position of 8 groups in the studio, using the distance to the classroom boundary as parameter to control the major circulation space from the entrance to each of the studios.
Input:
a. Coordinates of 8 rectangles’ centroid
b. Rectangles’ size
c. Boundary of classroom
Optimization process:
a. Minimize distance from classroom boundary to the rectangles’ centroid
b. Minimize overlapping areas between rectangles
Step 2 Each group contains around 12 people(11–13), a set of circles(representing movement range of different types of desk for discussion and individual work) are distributed in each group boundary(the rectangle).
For each group, the circles are optimized to repel each other (privacy between students) while attracting by the rectangle (collectiveness within the studio). And between groups, the circles are also set to repel each other. Both privacy and collectiveness are controlled by the overlapping area between circles and rectangles. The process is conducted group by group instead of 8 groups together.
Input:
a. The number of different types of desk
b. Coordinates of circles’ centroid
c. Circles’ radius depends on types of desk
d. Rectangles from last step
Optimization process:
a. Optimize the number of each desk type
b. Minimize the overlapping areas between circles
c. Maximize the overlapping areas between the circles and rectangles
Step 3 Manually operating and adjusting the orientation of the desks to better fit in the classroom for circulation
Results
W e try to optimize the position of the desks all together, however, the result doesn’t show any trend (figure.16). Therefore we decided to change our plan to optimize group by group, so that the trends could be salient. As a result, from the Discover chart we could see that the trend is salient in all the optimization (step1 & 2)(figure.14&15).
In step 3, we also try to rotate the desks in certain angles as input parameters, and set the constraint that the overlapping area should be less than 0. However, Discover doesn’t give us any optimal result, so we choose to manually adjust the angles of each of the desks.
After all the research and experiment we’ve done, we use the procedure introduced above to get to our final design of the re-organizationg of free studio seating. Here are some of the drawings and diagrams:
Conclusion
During the whole process of experimenting on Grasshopper and Discover optimization, we found that although it is really convenient and effective when computation starts to engage into the design process, there still exist problems. Firstly, the harsh constraints will be obstacles so that no suitable result could be generated. To solve this, we need to manually control the input in order to fit in the constraints. Secondly, there might not be a trend in the optimization when too many inputs and objectives are set, for example, the optimization process of desks position all together. Thirdly, we still need to manually adjust the position and angle of desks so that it would be more appropriate for the activities happening in the studio.
Generally speaking, the final design indeed improve and balance the privacy and collectiveness between students, to provide possilibity to generate ideas and enhance communication.
Overall, the idea of using computational design tools, like Grasshopper, Python, and Discover, to conduct an optimizing process is feasible. It helps a lot in reducing human and time costs, generating multiple solutions towards demand. However, today’s computation still needs manual operation to achieve the best and appropriate result. On one hand, the future development direction could be finding ways to make the computation process run more accurately and intelligently, to achieve the designer’s purpose and avoid semi-manual operation. Or, on the other hand, to maintain a better balance between computational design and manual design by hand and mind. ■
