CDfAM 2024 wrapped up last week, and my head is on fire with thoughts and inspiration. Here is my take on the worlds best conference for computational design.

There is no community like ours.

How many other communities can boast about having a rockstar figurehead like Duann Scott throwing events in such inspiring locations? There was nowhere else I would have rather been than Berlin’s AXICA Convention Centre, together with our group of makers, trailblazers, academics, and industry giants to share in what we all love.

We truly had it all. From legends such as Ole Sigmund and Andreas Vlahinos to big hitters like Dassault Systèmes and architects who truly blew me away. Nothing was missed. The line-up had something for everyone, and we all learned something new.

I am so happy to have found myself in this community, and those two days have truly left their mark. Everyone I got to speak to was open, willing to share ideas, and approachable. I even got my fanboy moment when Bradley Rothenberg (nTop CEO) came and shook my hand saying he’d seen some stuff of mine.

There are no more excuses.

The computational design rule book is now clear to us all. You can no longer get away with ignoring the manufacturing process. If you are not using optimisation, you are doing it wrong.

I for one needed to hear this.

Simulation is advancing, and if you aren’t up to speed you will be left behind. Traditional simulation methods are not going anywhere, but meshless (yes, I said it) is here to stay. You can’t shy away from mechanical performance because you decided to overcomplicate your design. “It’s too difficult to predict because there’s a graded gyroid inside” is not going to cut it anymore. Hyperganic and Metafold get it.

And you have to take the manufacturing process into account. Leaving overhangs, thin walls, and other nasties in your designs won’t do. We’re computational designers (whatever that means) so get your computer to do it for you.

Printers are not perfect, so stop expecting them to be. Take uncertainty into account, make robust parts.

AI came to the party but showed up late.

The AM community loves hype. We are the first ones to jump on every bandwagon, and we are guilty of overselling technology. So I was pleasantly surprised to see how we are dealing with AI.

We’re still not sure what terminology to use, but on the whole, we are getting more mature. We also have Ole Sigmund, Andreas Vlahinos, and Kristen Edwards from MIT to keep our expectations in check, and I think we are being quite sensible.

That’s not to say we aren’t innovating. Navasto gave us a glimpse into what the future of AI-accelerated CFD analysis could look like. But if Andreas says “It’s too good to be true!”, I’m inclined to listen and wait to be truly convinced. But this reality seems tantalisingly close.

I, like everyone else, still want to see progress in using generative AI for full-on object creation, but the lack of representation on the stage may tell us everything we need to know.

For now.

Are we really democratising manufacturing?

“3D printing Additive manufacturing democratises manufacturing”

The buzzwords we hear all the time. And it’s sort of true? You can buy FDM and SLA printers for next to nothing these days.

But it’s also not quite true.

Charging thousands to huge aerospace companies for your niche solution probably feels like success, but maybe you’re putting cutting-edge design tech in the hands of the wrong people.

Make high-performing software accessible to hobbyist computational designers like myself to play with (please), but also to small start-ups like Metamorphic, who have a track record in providing exceptional solutions to the most complicated design challenges, with lower-cost software options. These people love their trade so much that they grind through the process of starting their own businesses. Imagine where we could get to if their passion and your tech were brought together.

The open-source community still had strong representation with Aaron Porterfield, the creator of Crystallon. His love for making and breaking stuff resonated with me, as well as the laid-back presentation which stood out from the salesy stuff.

And you know you’re doing something right when some of the most talented designers in the world use your code.

We need to talk to architects more.

One of my favourite aspects of CDfAM was the opportunity to see the incredible work of some of the world’s best architects, and nerd out about design software at the end of the day over a beer. It was so cool to talk about design and ask questions from people with completely different perspectives.

I always had an appreciation for architecture, but I never fully appreciated the whole process until now.

Calculating the glare from glass panels into pilots eyeballs.

Cable management optimisation.

Wifi signal attenuation modelling.

Airports made out of thousands of unique, curved wooden sections.

Us AM designers don’t know what data-driven design means in comparison. We should be talking to the experts on how they manage data for these large-scale projects.

This is [not] computational design.

So after all this, we still don’t know what computational design is.

Patrick Pradel from the UK DfAM network showed us how diverse we are, all coming together from such a wide array of backgrounds. He also asked us how we should be training future computational designers, and whether our current education is up to scratch.

You might be tempted to think that the lack of formal CDfAM education is a problem. But sitting there in a room full of creative folk from biotech, marine structural engineering, computer graphics, product design, and so many other walks of life, I was full of confidence.

We haven’t done so badly up to this point in time. Our community is special because we are all different but doing something we all love.

We create robust designs by considering not only what we already know, but also stochastic effects we don’t know about. We shouldn’t encourage future members of our community to channel their views with formal education. Instead, we need to embrace their diversity and include the uncertainty.

Put the “design” in CDfAM

The closing act, Onur Yüce Gün from New Balance, could not have done it better. We had gotten a little lost in hardcore engineering, simulation and science over the two days but Onur slapped us in the face with thought-provoking creativity and abstract ideas.

We are a little divided in whether we should be removing humans from design and automating everything, or enhancing designers with better tools. Onur left us all implanted with the idea that design and humans cannot be separated.

Without humans, what is design?

Oh, and we still have no idea what computational designers are. But I kinda want to keep it that way.