Volvo Employs AI To Redefine Car Safety

Volvo Cars is using AI-generated virtual settings and real-world incident data from advanced automobile sensors to improve the development of their safety software, such as driver assistance systems (ADAS).

Volvo hopes to prevent similar events in the future by recreating and analysing them in innovative methods such as employing Gaussian splatting to generate realistic 3D scenes from real-world photos.
Gaussian splatting allows the system to manipulate virtual environments, enabling the alteration of road users and traffic behaviors to generate diverse outcomes. This allows for the exposure of safety software to a wide range of traffic conditions at an unprecedented speed and scale. Consequently, the automaker can now design software that effectively addresses complex and unusual 'edge scenarios,' reducing the time required for exposure from months to mere days. Alwin Bakkenes, head of global software engineering at Volvo Cars, said: “We already have millions of data points of moments that never happened that we use to develop our software “Thanks to Gaussian splatting, we can select one of the rare edge 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.”

The firm has a long history of using data to improve safety. In the 1970s, Volvo's Safety Research Team manually gathered data from crash sites, which led to the development of life-saving technologies like the Side Impact Protection System and Whiplash Injury Protection System.

Currently, Volvo is collaborating with Zenseact to create a virtual environment for safety advancements.
The development of AI-driven safety software is supported by Volvo Cars’ expanded partnership with NVIDIA. NVIDIA's accelerated computing and sensor-driven AI analysis will be integrated into Volvo's latest generation of fully electric cars to enhance safety technology.

The data will be contextualized, new insights will be unlocked, and future safety models will be trained on an AI supercomputing platform powered by NVIDIA DGX systems.