The Evolving Architect's Toolkit

an architect sitting at his workstation working on Revit on his computer
  • Blog
Nick Carter
Tags:
  • Architecture

Fear the Future?

Computer-aided design: CAD. Three words that shook our industry a little more than 50 years ago. ‘What will happen to our jobs?’, ‘But, they don’t have the skills and knowledge I have!’ and ‘They’ll never replace me.’ Three phrases that were said then, but recently we’ve been hearing these again. We are hearing them around conversations with automated, parametric, and generative design. Uncertainty brings fear, so let’s clear this up.

Black and white photo of architects drafting by hand

Automation

Automation is already very common in many industries including our own. We simply don’t notice this because it has become the status quo. It exists within CAD, with tech-savvy architects programmatically telling the computer step-by-step where and how to place things. Let’s use parking spaces as an example. Being able to quickly create a parking layout gives valuable insight to how certain designs perform but placing those parking spots manually may not be of any benefit to the architect - the results are what they are after. This is the perfect job for automation as it allows architects to spend time on more important tasks. Automation exists in our new 3D world in many of the same ways, but automated design provides more predictability thanks to the data within these elements/objects. With this information there is greater understanding of the steps we are programmatically taking.

step by step how to do Automatic Parking Count in Revit

Parametric Design

These models we build and their elements contain a lot of data. Parameters include Height, Length, and Location while others contain information about the Manufacturers and the Costs. Some of these parameters obviously dictate how the element looks, such as the height. We can see the height in 3D. What parametric design does is create relationships across these elements and their data. We link together their parameters to dictate different conditions to each other. Think of a window and all of the different configurations that it could have. Having the ability to automatically create shading, or even dictate suite layouts based on those configurations is a great advantage to designers. What’s important to take away from this is that it’s not that the designer can’t do this. They have already set the rules of what the design should do. They can tweak and change those rules, but fundamentally they should all react to their own conditions that they are tied to.

24 iterations of the same massing of a tower and podium on an irregular shaped site

Generative Design

When the powers of Parametric and Automated Design combine forces with Machine Learning, we get something we can call Generative Design. What is machine learning? It’s basically giving the computer precedents, defining what is important in those precedents, and then asking it to produce something based on what it has just learned. These outputs can be a hit, but most often they are a miss. Using Generative Design, we can assign parameters that we think are important. The computer is then able to quickly and automatically come up with many different design options. As it does this it reports back its parameters, and we are able to quickly see graphically, what has ‘hit’ and what has ‘missed’. From there we can narrow down which designs meet our criteria; we can see what works, how it works and further develop our design intent. This works great for things like Master Plans, where we use massing to fully understand how different sites work. This becomes a great design process as we can see options that we may not have otherwise and shape the direction those can go.

There is also another powerful way to implement Generative Design, and that’s as an ever-adapting ‘living thing’ that appears to design itself on the go. Instead of waiting for results like in the previous example, the results in this case are instant. Imagine being able to place the furniture layout of a room or suite based on the unit’s size and shape. Those furniture elements seem know how to adapt to the room and the elements around them. This is real-time generative design – it’s not as simple as it may seem when we see it in action. Each element must be given a set of rules to follow, know when conditions are met, and communicate those conditions back to everything else in the design. It must follow rules about accessible design, travel distances or even the light and shadow impact of a given space. It is easy to imagine all the bad possibilities that this can create, but that’s why it is important for architects to define what is important in their design – and have that in such a way that we can see how decisions are made.

colourful 3D rendering of a condo building made using BIM software

The Future is Friendly

It’s important to understand that we are not simply letting the computer design for us – it’s following our rules that we have set up. Those rules are tied to a lot of decisions about how we, as architects, want our spaces to perform. It gives us insight, and unrealized decisions to how to approach a design. If we imagine how we once viewed CAD as a job killer, today it’s viewed a job essential for any Architect. We believe these tools will become so commonplace in our industry that it will be hard to imagine a world without them. The ones who understand and can utilize these currently cutting edge concepts today will be in a great position when they inevitably become the industry standard.