Ethan Sun

Number of units: 2

image

Step 1 - Generative Design Framework

A very brief description of the design decisions from Step 1 following the Generative Design Framework.

  • Number of stories in a building (maximize value and comfort, minimize cost)
    • Design Variables
      • Slab Depth: The smaller the slab depth, the more stories can be built, but the higher the costs.
      • Number of stories desired.
    • Constants
      • Building Height: This is typically limited by the building code and zoning regulations. The developer is assumed to choose the maximum height allowed.
    • Evaluators
      • Clear span between floors as a proxy of comfort: The smaller the clear span, the more stories can be built, but the lower the comfort of the residents. A too large clear span is also not desirable, so maximum comfort will be in a certain range.
      • Value of building: The sum of the areas of each floor weighted by height.
      • Cost: A function that is the inverse of the slab thickness.
    • Most Important Tradeoffs to Consider
      • Comfort vs. Cost
      • Comfort vs. Value
      • Value vs. Cost
  • Evacuation efficiency (minimize evacuation time and construction cost)
    • Design Variables
      • Placement (x, y coordinates) of staircases in a large shopping complex.
      • Number of staircases.
    • Evaluators
      • Distance to staircase: E.g. Average of the 10 largest minimum distances from a store to a staircase
      • Cost: Number of staircases * fixed amount of cost
      • Space Usage: Penalties (additional costs) for the staircase blocking valuable retail locations, such as cinemas, showrooms, food courts.
    • Constants
      • Floor Plan: Assume it is predetermined.
    • Most Important Tradeoffs to Consider
      • Cost vs. Evacuation Time
  • Placement of columns in a parking structure (maximize efficiency and minimize cost)
    • Design Variables
      • Spacing of columns
      • Width of parking spot (e.g. provide accessible parking)
    • Evaluators
      • Total number of parking spots (spots cannot overlap with a column)
      • Percentage of wasted area
      • Construction cost as a function of spacing between columns
    • Constants
      • Floor Plan
      • Driveways
    • Most Important Tradeoffs to Consider
      • Space usage vs. construction cost

Step 2 - Generative Design Study

I’ve chosen to make a generative design study on evacuation efficiency. Let’s say that we are given an arbitrary floor plan of a large shopping complex, and we are to determine the best number and location of staircases to place on the floor plan. Our objective is to maximize the evacuation efficiency but also minimize the potential costs of installing staircases.

I modelled this problem through creating the floor plan in Revit with multiple rooms representing the different stores in a shopping complex. Then I placed three (can be more) elements arbitrarily on the floor plan, which represents the location of the staircases. I computed the Manhattan distances from the location of the stores each of the three staircases, and the minimum of distance from a store to a staircase would be its evacuation time.

The design variables for the Generative Design study would be the number of staircases desired and the location (x, y coordinates) of each staircase. It is expected that evacuation efficiency would increase, but costs would also increase when the number of staircases increases. The location of each staircase may block desirable locations for placing profitable storefronts (such as cinemas and showrooms), so they have their associated costs (although this impact is currently not measured).

The constants is the floor plan in the Revit model. This is assumed not to change, since a different floor plan certainly yields different locations for optimal staircase placements.

The evaluators are cost and evacuation time. Evacuation time (Distance Indicator) is measured as the average of the 5 highest evacuation distances. Cost is a function of the number of staircases and the maximum number of people evacuating to a single staircase. The latter is taken into account because stairs that accommodate a lot of people must be designed to a higher standard. Having a even distribution of people going to different stairs also increase evacuation efficiency.

The below screenshot shows the floor plan of my study.

image

Step 3 - Generative Design Study Results

The Inputs and Outputs are shown below.

image

The Explore Outcomes page is shown below.

image

Scatterplot illustrating the tradeoff between cost and evacuation time.

image

From the scatterplot, we can see that 1-staircase generates the case with the lowest cost, while 3-staircase generates the case with the fastest evacuation time (shortest distance). When there are 2 staircases, many cases are provided, showing that costs increase when distances decrease. The overall behavior of the generative study is just as expected, showing a clear tradeoff between cost and distance. When there is 1 or 3 staircases, there is only one optimal set of locations because it is only performing single objective optimization (either cost or distance) since the other objective is not competitive with the 2-staircase outcomes.

In this case, the designer would likely pick a intermediate case to balance between cost and distance. It would likely be chosen from one of the 2-staircase outcomes. The takeaway is that the designer would have to sacrifice one objective to perform better in the other.

1-staircase outcome:

image

One of the 2-staircase outcomes:

image

3-staircase outcome:

image

Dynamo Graph

image