Polestar turning to AI to shape future models

Artificial intelligence is infiltrating the car world and could not only help shape future models but also get them on the road sooner.

That’s the bold call from Polestar design chief Maximilian Missoni, who was speaking at the international media launch of the upcoming Polestar 3 and 4 about the influence of fast-moving machine learning in the automotive space.

Missoni said his Swedish-based team was already dabbling in AI in styling future vehicles as a means to speed up the design process and broaden the options when creating a new car.

“AI will be playing a bigger role everywhere,” said Missoni, clearly excited by the potential of the new technology but equally wary of the challenges.

“The guys are using it to enhance their sketches. There’s software that you can put a sketch in that it colours it and gives it variations. You can give it a colour theme or a photo that you like or a theme or atmosphere you like.”

Missoni said AI still relied heavily on human input but that it could be leveraged for the benefit of the broader design team.

“It’s very different outcomes as to who does it,” he said, saying that the results that a computer could generate were reliant on the input – effectively what a human instructed it to do.

“Different designers have a very different outcome using the same tools,” said Missoni.

He said the early stages of leveraging AI were not about letting a computer design a car from scratch but taking some of the grunt work out of styling a new model.

“It’s not really to create ideas but it’s to help them speed up the process.”

While the large 3 and mid-size 4 SUVs are the hot news from Polestar currently – each is due on Australian roads within months – it’s the upcoming Polestar 5 and 6 that are more interesting from a design perspective.

Each sits on a new bonded aluminium architecture rather than leveraging other platforms from within the Geely Group, as the Polestar 3 and 4 do.

The Polestar 5 will be a Porsche Taycan-rivalling five-door liftback with an emphasis on performance and driving excitement.

The Polestar 6 will be an all-electric two-door roadster positioned as the performance pinnacle for the brand.

Unsurprisingly, there is some trepidation internally as designers learn what AI can do. Allowing computers to radically alter the original design themes and ideas was meeting internal resistance, for example.

“The team is a bit reluctant to do that. I think they feel that is not an original idea, and I get that,” said Missoni.

“It changes the value of the idea when you know it’s been created by a generated algorithm.”

And Polestar’s design chief is adamant that despite the potential to change the workflow and way of doing things, humans were still a crucial part of the car designing process.

“[AI] gives you options, but you still have to decide yourself. You can’t replace the decision maker. I really want my designers to take a lot of decisions themselves,” he said.

Missoni added that one of the big advantages of AI was shortening the development time in bringing a vehicle from concept to showrooms.

“Technology in electrification is moving so fast that you need to bring out models quicker,” he said, adding that the usual four-plus-year development cycles were now being shortened to less than three years in some instances.

But he warned that some parts of the development process still required time and human effort to ensure the vehicles were up to consumer expectations.

“That part of the process… validation, testing… you really shouldn’t take any shortcuts,” said Missoni.

For everything you auto know about EVs, listen to carsales’ Watts Under the Bonnet: the electric car podcast