Volvo has begun using artificial intelligence (AI) to train its driver assistance and safety software, as it continues to invest in AI features.
Its Gaussian splatting technique can simulate road conditions in a three-dimensional virtual environment based on data collected by its cars’ sensors, and was developed with Volvo’s AI arm Zenseact.
“We already have millions of data points of moments that never happened that we use to develop our software,” said Volvo head of global software engineering Alwin Bakkenes.
“Thanks to Gaussian splatting we can select one of the rare corner cases and explode it into thousands of new variations of the scenario to train and validate our models against. This has the potential to unlock a scale that we’ve never had before and even to catch edge cases before they happen in the real world.”
Volvo expanded its partnership with Nvidia in September, saying its new EX90 model would be powered by an Nvidia Drive Orin chip.
The AI supercomputing platform used to analyse these cars’ sensor data is powered by Nvidia’s DGX systems, Volvo said.
The company released its first AI-generated advertisement earlier this month, targeting Saudi Arabian buyers as Volvo hopes to reestablish its presence in the country. The video was made with Midjourney, and does not include images of any Volvo models.
Porsche also said it would use AI to train its assisted driving software last month. Porsche’s technology similarly collects data from its vehicles’ sensors to determine the likelihood of certain road events.
Volvo Australia said last week that it would delay its plans to only sell electric vehicles in Australia from 2026, after the global company said it would revise its goal of being electric-only by 2030. Volvo will instead aim for 90-100% electric sales by the end of the decade.
Volvo’s (STO: VOLV-B) share price closed at 308.90 kr, down from its previous close at 312.30 kr. Its market capitalisation is 628.09 billion kr.
Related content
