Hello Professionals 
,
A few weeks ago, I started exploring and trying to deepen my understanding of Generative Design for Autodesk Revit. I’ve managed to get a good grasp of writing basic scripts in Dynamo, but I’m still struggling to fully understand and create Generative Design scripts.
I’m very interested in learning about the logical steps to get started with Generative Design. My ultimate goal is to achieve something similar to the masterclass by @jacob.small titled “How Generative Design Can Help Optimize Neighborhoods” on YouTube, although I realize I’m not there yet. Unfortunately, I’ve found it challenging to locate tutorials or information about Generative Design online.
Any advice or guidance you could offer would be greatly appreciated.
1 Like
Do you have a Generative Design Problem?
One of the biggest mistakes you can make is trying to apply GD where it’s not needed. Finding a GD Problem and defining its design characteristics, inputs, outputs, constraints, conditions, etc. is usually the biggest and hardest task in proper GD development. There are plenty of small example tasks that you can automate through GD and eventually scale into actual design tools, but you want to make sure you have a valid use case first.
If you currently have some ideas for the types of problems you want to solve with GD then let’s start there. There are many different “types” of GD problems and you’ll likely handle each of them differently. Knowing and understanding your GD tool starts with knowing and understanding the design space with which you’re working. The graph development is really one of the last things you do.
2 Likes
For just learning ‘how’ I presented my favorite ‘my first generative design graph’ to the Chinamo User Group a few years back. The recording here: https://www.youtube.com/watch?v=42SylTdN3CY
As @Nick_Boyts mentioned the key part is understanding what a Generative Design use case looks like; not everything is - if there is a ‘single solution’ which you can work out mathematically then you’re using the wrong tech. But if you’re facing a wicked problem then GD is ideal, and design is a wicked problem in general (though not all aspects thereof). Once you confirm there is a wicked problem, then you can move into a prep step to build the logic and evaluation methods on paper, and then you can finally start Dynamo. For each hour of graph/code authoring I find I have to spend the same time on prep. Even when you ‘know exactly how’ you would do the design of a thing you often skip the finer points of the step - so build out that logic carefully, as if you skip something you might set yourself back to step 0 and going back to a blank canvas isn’t very fun.
2 Likes