Modular Placement and Usability Review (Generative Design) _ Part 2

 

https://www.linkedin.com/pulse/modular-placement-usability-review-generative-design-wk4zc

E(ent) entrance, T(1x2), T(1x1.5), R placement and shortest distance generation

The shortest distance was output to confirm that the bathroom's position was as far as possible. However, with L, K, M, and F not yet placed, it was confirmed that many variables exist to determine an appropriate placement.

As a result of confirming the furthest L value, the justification for R's placement was also found to be insufficient. After placing K and L, additional review of R's placement suitability is needed.

Even with a close L value, it is possible to derive the conclusion that a long distance can be calculated in reverse depending on the placement of R, F, and M.

Finding Usable Space Through Additional Placement of F, M

Confirm GD results to consider L, K placement in the remaining space after placing E, R, T1x2, T1x1.5, M, and F.

If the bathroom distance is short, a layout where circulation paths detour due to the placement of other spaces was expected, but the desired value was difficult to find.

If the bathroom distance is long, it is generally difficult to find highly usable L, K placements.

Review is needed for a method where L, K space 2x3 (6 units) is placed and connected to the entrance first, then the remaining space is placed in the remaining areas.

L, K Space Placement and Result Confirmation

Review the method of placing L, K space 2x3 (6 units) and the connection path to the entrance first, then placing the remaining spaces in the leftover areas.

After placing L, K, the remaining spaces find their way in, but empty spaces often occur at corners.

The same applies when the bathroom distance is long; corners are often not found.

Review is needed for a method of placing T or F, M first in the corner spaces, then placing R in the remaining corners.

T, F, M Placement Priority

After E, L, K placement, T 1x2, T1x1.5, F placement, and then R placement, it is difficult to find a meaningful cluster placement similarity for R.

Therefore, R is placed last to induce a meaningful spatial cluster for R in the empty space.

Additionally, the placement priority of T, F, M is set to the exterior first, and R's cluster placement is induced last.

Generative Design TEST [Result review based on Seed coefficient change]

Meaningful layouts confirmed in 10 out of 200 random seed results. Since results close to meaningful layouts appear with a probability of 5/100, the test was conducted by increasing the seed to 500.

Among the 500 seed results, an error occurred where the external placement count exceeded 60. The reason is that after E, L, K, T1x2 placement, an area where F, T1.5 cannot be placed occurs. (F, T1.5 are 1x3 long forms, so in the above results, a long form of 3 units cannot be placed. 1.5 spaces can be placed separately, but a change may be needed depending on meaningful space utilization.)

Filtering the previous error with an output that checks the number of mass placements yielded 54/500 results. The probability is similar to the previous one, and meaningful placements can be confirmed, but impossible placements are also confirmed. A method for filtering impossible placements is needed.

Review of External Placement Results (1)

External Placement 0 [34/50]

Accessible spaces are green; inaccessible spaces are red. Green spaces are considered usable for remaining/ancillary spaces, but isolated spaces can occur.

solated spaces may be usable via separate paths, but detailed planning is needed.

Additional review of external placement results for 1, 2, and 3 will be conducted with 24 results each.

Review of External Placement Results (2)

External Placement 1 [7/24]: Useful spaces found with <30% probability; needed to complement External Placement 0.

External Placement 2 [3/24]: Space utilization significantly drops to 1/8 probability; requires further review.

External Placement 3 [0/24]: 0% probability; increasing population size is needed, but further external placement values seem meaningless

Conclusion of This Test Round: Visible results for meaningful space placement are possible with L, K placement and external placement constraints. Additional rules, logic (e.g., Room cohesion, passage space, ancillary space beyond rec, door position, extra constraints), and 2D/3D output of the shape are needed.

Dynamo Script Overview

1. 5x5 placement points

2. space rec creation

3. geometry checks

4. sequential space rec placement

5. external placement

6. shortest distance

7. Mass

1. 5x5 Placement Points

Points and boundary lines are created for a 5x5 space, centered at (0,0,0).

Initial E (ent) entrance forms a red box, assumed movable within external boundary (excluding outer frame) for future L, K placements.

2. Space Rec Creation

Maximum space from CAD is created as a rectangle (rec) and sequentially placed.

Rec shapes are initially superimposed at (0,0,0) and moved to their positions by sequential placement (step 4).

3. Space Frame and Placement Space Geometry Check

A space frame is created to check for overlaps before sequential placement.

Doesintersect node filters spaces escaping the exterior, and placement seed values select points for each rectangle.

Sequentially placed forms are recorded as solids and shaped by GD.

4. Sequential Space Rec Placement

Placement proceeds in the order: [E, L, K > T1x2 > T1x1.5 + F > M > R].

This sequence and additional placement logic will likely change for utility and optimization.

5. External Placement

Solid is created in the interior space for 5x5 external boundary placement as previously described.

Overlapping space placement points form a Count value, and their sum is the External Placement Output.

6. Shortest Distance

Compares distances between bathrooms (T1x2 : T1x1.5), treating spaces other than E,L,K as obstacles.

Results can be filtered for distant placement, but will require refinement after incorporating door positions and detailed obstacle information.

7. Mass

Outputs the final number of placed masses. This output captures the last error of this TEST, as exceptions like the image on the right occur.

If Mass 10 outputs all values, Mass 3 outputs an error like the image above.

LK Closest Door Placement

Geometry.Intersect extracts possible door placement curves by checking overlaps between the previously calculated passage space solid and outer curves.

The distance from these curves to the LK Solid's centroid is measured, and the shortest curve is passed to the next group.

All walls are checked for overlap with the identified curve to find wall element hosts.

Doors are then placed using Familyinstance.Byhostandpoint, with the overlapping wall as the host, the extracted center point as the location, and the selected door family type.