Design decisions are rarely simple -- they typically involve evaluating many options and thinking about the tradeoffs between the features offered by each option.
In this assignment, you’ll create a simple Dynamo geometric model of a design decision that you’d like to explore:
- providing inputs for the design variables that you can choose between
- reporting key measures of the design that can be used to evaluate them
Then, you’ll use the Generative Design tool in Revit and Dynamo to:
- generate a variety of options to be evaluated
- consider the tradeoffs between the features offered by each option
Generative Design Framework
As you get started on this assignment, take off your programming hat and put on your design thinking hat. The most important step in this assignment is thinking about what is the problem you want to solve or the question you want to answer. Once you draft your problem on paper, think how to frame a design decision in a way that lends itself to an optimization approach. In a nutshell, the Generative Design methodology can be broken down into:
- Objective: What is the question you are trying to answer with the Generative Design study? For example: What is the “ideal” shape of my building to maximize sun exposure and have minimum facade area? The more specific the better! We will refine this Objective statement once we define the following pieces of the framework, but usually we want to maximize or minimize an output from the model.
- Model: How would a model to solve the problem (in its simplest form) look like? For the building shape study, we need to model the building using a base geometry, then lofting between cross sections and then converting into a solid, for example. This part of the framework should list the inputs (Design Variables and Constants) and how success looks like in terms of outputs. For this example, the inputs are base cross-section height and width, z offset, top cross-section height and width and rotation angle between the two cross-sections.
- Design Variables: These are input numbers that drive your model. By tweaking these, you will explore different generations of the model. For our example, we want to focus only on the top cross-section height and width, and rotation angle.
- Constants: Inputs that don’t change and remain the same for the entire generation set. For this case, the base cross-section parameters and z offset remain unchanged. Note that even though these values don’t change, they are needed to build the lofted surface of our building shape.
- Evaluators: These are output metrics that define a good or bad performing design. Evaluators should always be calculated from Design Variables and Constants. For all Generative Design studies, the outputs must be drawn from the objective. In our example, we want to maximize sun exposure which can be measured from average hours of direct sunlight, and minimize facade area which can be measured from calculating the surface area of our building shape in Dynamo.
- Interpretation: Generative Design in a systematic way to explore multiple design options that fit into a constrained model. However, it’s important to interpret the results and understand the tradeoffs between different variables. Just like in a system of equations, there can be multiple solutions, one, or none. This is also true in Generative Design and actually it means that one solution (even partial solution) can lead to tweaking the rules of your model to explore a new branch of design options. What behavior can you see from the generations? Can you rule out some designs or are there generations that don't make sense?
Step 1 – To warm up, describe 3 design decisions from any discipline using the Generative Design Framework outlined above. For this step you don’t need to model (yet), it’s just to get familiar with explaining a design decision using the framework
- Start by thinking creatively about the design decisions that are typically required on projects in your discipline (for example, structural, construction, architectural, sustainability) and which of those decisions have the biggest impact on the success of the project.
- Choose three of those design decisions and briefly outline (for example, using bullet lists):
- The most impactful design variables -- values that must be chosen that have the biggest impact on the quality of the outcome.
- The most important evaluators -- values that can be measured to score and evaluate the essential tradeoffs between the outcomes resulting from the design variable choices.
- The most important tradeoffs that are typically considered in making that design decision. Every design decision has its own tradeoffs, so think creatively about how to describe what’s truly important. Some examples might include:
- Construction Planning
- Construction Time vs. Construction Cost vs. Product Quality
- Fabrication Cost vs. Assembly Time
- Structural Design
- Member Weight vs. Deformation or Drift
- Bay Size vs. Materials Cost
- Architectural Design
- Materials Cost vs. Appearance Quality
- View Quality vs. Shading
- Capital Cost vs. Operating Cost
- Material Cost vs. Carbon Footprint
- Some examples of Design Problems that fit well with the Generative Design Framework include:
- Views in the City, having as input the u,v coordinates to place the building, total height, story height; and as constants the surrounding buildings. The output to maximize is the unobstructed views.
- Stadium seating configuration to maximize the people seated with views to the center of the pitch. Using distance to expectators, angle and number of rows as inputs and number of observers and volume used in the stadium as a proxy of cost.
- Orientation and dimensions (base and height) of tiles on a floor to minimize the trimming and waste of material.
- Spacing of members on a simple truss to maximize resistance up to a certain safe value, minimize deformation and minimize material used.
- Where to put the staging area on a construction site to minimize movement of material. Modeling site areas as rectangles and measuring distances as polylines.
- There’s no absolutely right or wrong answer here -- the point is to think creatively about how to frame the design decisions in a way that will lend themselves to an optimization That will inform your decision-making and have the best impact on the final outcome.
Step 2 – Choose one of those design decisions (from Step 1) and create a SIMPLE Dynamo model that can be studied using Generative Design. The Generative Design Study that comes from your design decision doesn’t need to be complex - this is a preparation towards Module 8 in which we will think more about the actual problem.
- Review the examples provided in this module for ideas about how to create simple Dynamo geometry and logic that can be used to model the essential features of the design decision. There’s a wealth of good work that you can learn from and adapt as you create your model -- use those examples to your advantage and don’t reinvent the wheel.
- Create a Dynamo model (also called a Study Graph) to explore the design decision in the Generative Design tool.
- For 2 units, create a Study Graph with:
- Two design variables (study inputs)
- Two evaluators (study outputs)
- For 3 units, create a Study Graph with:
- Three design variables (study inputs)
- Three evaluators (study outputs)
- For 4 units, create a Study Graph with:
- Four design variables (study inputs)
- Four evaluators (study outputs)
- As you design your study graph, think very carefully about the outputs that you are creating to evaluate the design options. They should be:
- Values that truly quantify the most important measures for evaluating the options relative to the tradeoffs being studied.
- Values that are relatively simple to compute -- for example, measures based on areas and volumes of geometry are OK. You can substitute more precise measures based on a more complex model at a later date. But as a quick starting point, simple geometric measures are often a good surrogate for understanding the impacts of varying the study inputs.
- Measures that pull the tradeoffs in different directions. You want to explore a tradeoff, so it’s important to capture measures that would lead you to choose different options:
- If the measures aren’t truly independent (for example, surfaceArea and materialCost, where materialCost = surfaceArea * constantValue), you won’t create a tradeoff between those measures. The Generative Design tool will solve for a single value.
- If the measure align (for example, roofArea and solarEnergyAvailable), you’ll typically create a linear relationship between those measures. While it’s good to understand that linear relationship, it’s not really a tradeoff.
- It’s often useful to focus on the measures that are most affected by the input variables. For example:
- If you maximize the glazing area on the wall of a building -- what measures are most positively affected? What measures are most adversely affected?
- Similarly, if you use maximize the column grid spacing of a structure -- what measures are positively affected by the longer spans? What measures are negatively affected by those longer spans?
Step 3 – Create and run a Generative Design Study using your Study Graph to illustrate the essential tradeoff in your design decision
- Create a new Generative Design Study
- Choose the inputs to be tested
- Choose the outputs that will drive the optimization and whether these outputs should be maximized, minimized, or ignored
- Generate the design options
- Visualize the results
- Use the Explore Outcomes reporting features to create a visualization that best illustrates the essential tradeoff that would inform your design decision.
- Set up a Scatter Plot or Parallel Coordinates graph that shows how the options generated by Generative Design perform on the output measures that you feel are most important.
- Capture a screenshot of that graphic.
- Provide a brief explanation of what’s being shown in the Scatterplot and how the tradeoff being illustrated would impact the design decision. What would you do with this info?
- Please create a folder named “Module 7” within your personal folder in our Autodesk Construction Cloud project:
- Then, upload these items to your Module 7 folder using the web interface:
- Your Dynamo Study Graph (.DYN) file
- Any supporting custom nodes that you used or created
- A brief document outlining:
- The three design decisions considered using the Optimization Framework (from Step 1).
- The screenshot of the Scatterplot illustrating the tradeoff that you chose to model and study.
- Your explanation of what’s shown in the Scatterplot and how it would impact the design decision that you chose.
- Create a link to your Module 7 folder:
- Right-click on the Module 7 folder in the file tree (at the left side of the interface) and choose Share from the pull-down menu.
- Choose Share with Project Members, then switch to the Link tab.
- Click the Copy button to copy the link to your clipboard.
- Create a new posting on this Notion page — Design Journal Entry: Study Your Options — including:
- A very brief description of the design decisions from Step 1 following the Generative Design Framework.
- A more detailed description of the design decision from Step 2 that you decided to run a Generative Design Study with. Use the Framework structure to keep it consistent.
- The screenshot of the Scatterplot illustrating the tradeoff that you chose to model and study.
- An image of your Dynamo Study Graph (showing all your nodes and the connecting logic) -- You can use the File > Export Workspace As Image... command in Dynamo to save a PNG image to upload with your posting.
- The link to your Module 7 folder on Autodesk Construction Cloud.