Volvo is using AI to make its cars safer: How is it doing it?

Volvo has long been synonymous with automotive safety, and now the company is leveraging artificial intelligence to take its crash prevention systems to the next level. Thanks to a groundbreaking partnership with Nvidia, Volvo is using Gaussian splatting, a cutting-edge 3D rendering technique, to train its vehicles to avoid accidents more effectively than ever before.

How Volvo is improving safety with AI

Volvo is implementing Gaussian splatting to create ultra-realistic 3D scenes based on real-world driving data. By analyzing sensor data from its latest vehicles—including emergency braking, sharp steering, and manual interventions—the company can reconstruct and manipulate driving scenarios to improve its Advanced Driver Assistance Systems (ADAS).

Unlike traditional AI training methods, this approach allows Volvo engineers to generate thousands of variations of rare “edge cases,” ensuring the AI encounters and learns from high-risk situations before they happen on the road. This dramatically improves the system’s ability to predict and prevent accidents.

The role of Nvidia’s technology

Volvo’s collaboration with Nvidia is key to this advancement. The dual Nvidia AGX Orin system inside the Volvo EX90 and the upcoming ES90 provides supercomputing power capable of processing massive amounts of sensor data in real-time. This accelerates AI training, making autonomous and semi-autonomous driving more reliable and efficient.

By harnessing this technology, Volvo is moving closer to its vision of zero collisions. While the EX90 initially launched without active Lidar functionality, future updates—powered by Nvidia’s advanced chipsets—will unlock even more powerful safety features.