Volvo: AI-Powered Virtual World Transforms Safety Testing

Volvo has unveiled a groundbreaking approach to improving vehicle safety software by harnessing AI-generated 3D worlds.

By using synthesised incident data collected from advanced onboard sensors, Volvo is able to analyse, reconstruct and explore various traffic scenarios at an unprecedented speed and scale.

The cutting-edge approach is powered by an advanced computational technique known as Gaussian splatting. Underpinned by Nvidia’s AI supercomputing platform, the technology rapidly creates realistic, high-fidelity 3D scenes and subjects derived from real-world visuals.

Unlocking new possibilities in safety software

Volvo’s ability to simulate vast amounts of traffic data transforms how vehicle safety is developed and tested. 

Alwin Bakkenes, Head of Global Software Engineering at Volvo Cars, highlights the significance of this innovation: “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 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.”

What is Gaussian splatting?

Gaussian splatting is a revolutionary 3D rendering technique that uses tiny 3D Gaussians, or ‘splats,’ to represent a scene. Unlike traditional methods that rely on triangular meshes, this approach fills the space with Gaussians with specific properties such as position, size, orientation, colour, and transparency.

To create an accurate 3D representation, multiple images of a scene are captured from different viewpoints. The Gaussian splatting algorithm then optimises these properties, effectively learning to represent the environment with remarkable detail and fluidity. The method enables smooth and realistic visuals, even in complex scenarios involving reflections and transparency.

The technique is particularly valuable due to its speed and efficiency, making it ideal for applications such as virtual reality, augmented reality, and advanced vehicle safety systems in Volvo’s case.

Creating safer cars with AI

By leveraging data from its vehicles’ advanced sensors, Volvo can now apply Gaussian splatting to train and refine safety software such as advanced driver assistance systems (ADAS). The method allows Volvo to expose safety systems to a broad range of traffic scenarios, rapidly generating and testing different conditions.

AI-generated virtual environments can adjust to modify traffic behaviour, introduce new obstacles and explore various potential outcomes. The flexibility is crucial in developing safety software that can effectively handle rare and potentially hazardous edge cases.

In an official statement, Volvo elaborated on the benefits of the approach: “We now can develop software that works well also in complex, rare yet potentially dangerous ‘edge cases’ and reduce the time it takes to expose our software to edge cases, from months to days.”

A commitment to autonomous driving innovation

Volvo’s adoption of AI-driven 3D environments is part of a broader commitment to autonomous vehicle technology. The simulations complement real-world testing and play a crucial role in training, development and validation, providing a safe, scalable and cost-effective approach to improving vehicle safety.

Volvo Cars and automotive safety supplier Autoliv co-founded Zenseact, an AI and software firm that supports its investment in advanced safety solutions. Zenseact specialises in AI development for autonomous driving and ADAS technologies, aiming to enhance road safety and reduce accidents.

Volvo’s collaboration with Nvidia further reinforces its focus on cutting-edge automotive technology. It utilises Nvidia’s AI computing power to support its software-defined vehicle architecture, enabling next-generation safety, connectivity, and data management.

Volvo’s latest flagship model, the EX90, is built on a centralised core compute architecture developed in partnership with Nvidia. Capable of performing over 250 trillion operations per second (TOPS), the system orchestrates everything from autonomous driving features to AI-powered driver assistance technologies.

As part of its long-term vision, Volvo has announced plans to integrate Nvidia’s DRIVE Thor system—an advanced AI-powered central computing platform for autonomous vehicles. The collaboration aims to introduce vehicles equipped with continuously upgradable, AI-driven safety and driving assistance systems.

Jensen Huang, founder and CEO of Nvidia, has emphasised the rapid advancements in AI and accelerated computing: “Advances in accelerated computing and AI are moving at lightspeed. DRIVE Thor is the superhero of centralised compute, with lightning-fast performance to deliver continuously upgradable, safe and secure software-defined supercomputers on wheels.”

Advancing safety through AI

Volvo’s strategic use of AI-generated 3D worlds represents a significant leap forward in automotive safety. By leveraging Gaussian splatting, Nvidia’s supercomputing capabilities and in-house AI expertise, the company sets a new standard for vehicle safety development.

With an unwavering commitment to innovation, Volvo is enhancing its current safety systems and laying the foundation for the future of autonomous driving. As the automotive industry continues to evolve, Volvo’s AI-driven approach ensures that safety remains at the forefront of its technological advancements.

Stay informed and inspired with the EV Magazine community by signing up for our free weekly newsletter. Each week, receive in your inbox the latest industry news and in-depth insights into the world of electric vehicles, sustainability and cutting-edge innovations

Subscribe now, don’t miss the opportunity to stay ahead in the fast-evolving EV industry.

EV Magazine is a BizClik brand.