How Volvo's AI 3D Worlds are Enhancing Manufacturing Safety

Volvo has introduced a revolutionary method for improving vehicle safety systems through the use of AI-generated 3D environments. By leveraging synthesised incident data gathered from its advanced onboard sensors, the company can now analyse, reconstruct and simulate a wide range of traffic scenarios faster and more extensively than ever before.

This state-of-the-art technique is driven by Gaussian splatting, a sophisticated computational method supported by Nvidia’s powerful AI supercomputing platform. The technology allows for the rapid generation of realistic, high-quality 3D scenes and objects derived from real-world imagery.

Transforming safety software development

Volvo’s capacity to simulate vast amounts of traffic data is transforming the way vehicle safety is designed and tested.

This breakthrough enables the development of more effective and responsive safety systems, marking a significant advancement in automotive technology.

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.”

Understanding Gaussian splatting

Gaussian splatting is a cutting-edge 3D rendering method that replaces traditional mesh-based models with a cloud of tiny 3D Gaussians, referred to as ‘splats’, to depict a scene. Each splat has attributes like position, size, orientation, colour and transparency, allowing the scene to be represented in a more flexible and detailed way.

To build an accurate 3D environment, images of the scene are taken from various angles. The algorithm then refines the splats’ properties, learning to recreate the space with high precision and fluid motion. This results in highly realistic visuals, even in complex conditions with reflections or transparent surfaces.

The technique stands out for its speed and efficiency, making it particularly well-suited for uses like virtual and augmented reality—as well as advanced safety development in Volvo’s vehicles.

Using AI to build safer vehicles

Volvo is harnessing data from its vehicles’ sophisticated sensors to apply Gaussian splatting in the development of safety software like advanced driver assistance systems (ADAS). This technique allows safety features to be trained and refined using a wide variety of traffic situations, enabling fast creation and testing of different real-world scenarios.

Through AI-generated virtual environments, Volvo can simulate dynamic traffic behaviours, introduce unexpected obstacles and explore multiple outcomes. This adaptability is key to designing systems capable of responding to rare, high-risk situations that may not be easily replicated through traditional testing.

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.”

Paving the way for autonomous innovation

Volvo’s use of AI-powered 3D simulations reflects its wider ambition to lead in autonomous vehicle technology. These virtual environments work alongside physical testing, offering a safer, more flexible and cost-effective method for developing, training and validating advanced safety systems.

To strengthen this pursuit, Volvo Cars partnered with safety technology leader Autoliv to co-found Zenseact—an AI and software company dedicated to progressing autonomous driving and advanced driver assistance systems (ADAS). Zenseact focuses on using AI to improve road safety and help reduce accidents.

Volvo’s partnership with Nvidia highlights its drive to remain at the forefront of automotive innovation. By tapping into Nvidia’s powerful AI computing capabilities, Volvo supports a software-led vehicle architecture designed to enable future-ready safety systems, smarter connectivity and seamless data handling.

Its flagship model, the EX90, showcases this vision. Developed with Nvidia, it features a centralised computing core capable of executing over 250 trillion operations per second. This system manages everything from autonomous driving tools to intelligent driver assistance functions.

Looking ahead, Volvo plans to incorporate Nvidia’s DRIVE Thor platform—an advanced AI-based computing system tailored for autonomous vehicles. The aim is to deliver cars with safety and assistance features that can evolve over time through continuous software upgrades.

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.”

Driving safety forward with AI

Volvo’s innovative use of AI-generated 3D environments marks a major step forward in vehicle safety. Harnessing Gaussian splatting, Nvidia’s supercomputing power and its own AI know-how, the company is setting a fresh benchmark in safety development.

Guided by a strong dedication to innovation, Volvo is not only refining its existing safety systems but also preparing the groundwork for autonomous driving. As the motor industry moves ahead, Volvo’s AI-led strategy keeps safety firmly at the heart of its technological progress.

Explore the latest edition of Manufacturing Digital and be part of the conversation at our global conference series, Manufacturing LIVE.

Sign up to our weekly newsletter here.

Discover all our upcoming events and secure your tickets today.

Manufacturing Digital is a BizClik brand.