Computational superpowers are here: Are we ready to use AI in design?

Computational design already plays a major role in engineering design and architecture, but much less so in service and graphic design. Computational design offers significant opportunities to boost innovation potential in the Finnish context. FCAI held a webinar on emerging AI methods in computational design in April 2022 to survey breakthroughs in deep learning, optimization, reinforcement learning, and simulation intelligence. The event gathered close to 100 participants mostly from companies interested in new tools and methods.

Computational design will be needed not only in traditional areas of design but also in creating novel societal solutions for climate change and sustainability, for example. We are not yet fully harnessing the opportunities that computational design offers, says professor Antti Oulasvirta from Aalto University.

“There are staggering numbers showing the importance of design in producing innovations that are the basis of economic growth and innovation. For example, top innovation companies have produced 75% of their revenue from services and products that did not exist five years ago,” said Oulasvirta.

“Computational superpowers are here but they are not replacing designers. Designers who use AI will replace those who don’t. The superpowers are more like small tools that are sprinkled around the design process that help us to be more productive, smarter, innovative, and offer value to the end-user,” continued Oulasvirta.

He encourages designers to let algorithms solve hard problems and to find novel solutions that competitors cannot offer. “Most of the things taking place so far are related to generating ideas, prototyping and testing products. Forerunner designers use AI also for defining problems and understanding people.”

There are interesting new applications helping designers. Oulasvirta presented several opportunities related to detecting design flaws, evaluating beauty, creating codes from sketches, mining photos, simulating users’ behavior and optimizing visual displays.

“The webinar gave a good overview how computational design can be used,” says Heikki Ailisto, research professor from VTT. “There are already many companies in Finland benefiting from it in practice and some of them in a very advanced way. The deployment requires designers with high-level skills in machine learning, simulation methods and artificial intelligence. Many companies are still considering whether to use computational design in a systematic way. Hopefully through this webinar we managed to encourage them to move forward.”

The webinar surveyed emerging breakthroughs in deep learning, optimization, reinforcement learning, and simulation intelligence among others. The presentation materials and video recordings can be found on the event page.