A new partnership between Swedish tech company RemotiveLabs and Google Cloud aims to speed new vehicle development and improve quality before even one physical component is built.

The partnership melds RemotiveLabs’s visualization platform called RemotiveTopology, with Google Cloud’s Horizon, an open-source software development platform, the companies announced early Monday.

The combination will boost Google Cloud’s concept of “system of systems” where entire virtual vehicles, so-called “digital twins,” can be designed and tested collaboratively in the cloud by various development teams, explained Florian Haubner, industry architect lead at Google, in an interview conducted via email.

“Horizon accelerates iterations for developers and provides the necessary scale for agents,” wrote Haubner. “We see reductions in feature development cost in the magnitude of 40%, improved quality due to higher test coverage and up to 9x faster software testing in virtual. With RemotiveLabs we are now expanding the benefits of virtual development beyond Android towards the full vehicle.”

That’s accomplished through RemotiveTopology, which has been in use for several years, adopted by such companies as Volvo Cars, Ford, Horse Powertrain, HERE Technologies and Arm.

It will now expand Horizon’s capabilities, according to RemotiveLabs CEO and co-founder Per Sigurdson who explained in an interview, “we’re bringing the rest of the vehicle into the cloud, so they have something to integrate against. You will use that to build and deploy your functionality in Horizon, and then you would use RemotiveLabs integration to bring in the rest of the vehicle when you’d want to do testing.”

A key goal is enabling much of the software design and testing to occur virtually nearer to the start of overall vehicle development, a philosophy known as left-shifting. By doing so, many issues are discovered before costly physical components are built, according to Sigurdson.

This is especially valuable when dealing with electric and electronic systems, with Sigurdson explaining, “they have different suppliers and different versions of the signals and whatnot, and you want to kind of iron out all of these mismatches and potential misunderstandings at a really early stage.”

Virtual simulations have actually been around for at least five decades, but improving the ability to collaborate in the cloud, along with assistance from AI has been a real difference maker on several fronts according to David Heiny, co-founder and CEO of SimScale, a company which specializes in such simulations.

“At the end of the day, if an organization is able to out engineer to have a higher engineering velocity, they are able to meet the customer demand faster,” said Heiny, in an interview.

Indeed, achieving such “engineering velocity” is often a challenge for companies as complex as automakers where various functions are often handled in separate organizational silos, a set-up not conducive to speedy collaboration, pointed out Sigurdson.

So a key goal of this new partnership is to provide easily available technology to break down those silos.

“One silo is between the software and the hardware domain, because there’s different competences, different tooling, and different time perspectives as well, and we kind of merge them together,” said Sigurdson.

An ongoing partner for RemotiveLabs is technology company Arm, with which the firms have been working together for the last 10 to 12 months to build a better platform for creating vehicle virtual twins.

Arm provides what Barrie Mullins, senior director of Arm total design for physical AI described as the “compute foundation” for the virtual twin, working with its partners to create virtual platforms and a virtual system on a chip (vSoC) that accurately represent Arm-based silicon and development boards, she explained in an interview conducted by email.

The result is the ability for developers to start software development, integration, testing, and verification before physical silicon or hardware is available.

RemotiveLabs extends that virtual environment from the compute platform into the wider vehicle, adding the electric/electronic architecture, networks, signals, virtual electronic control units and software interfaces, according to Mullins.

It’s another element in achieving that shift-left functionality.

“Once the product architecture or form factor is defined, developers gain access to continuous, representative workloads without having to wait for physical hardware,” said Mullins. “Software integration, system testing, and AI model validation can start much earlier using realistic scenarios. Work that previously required development boards, test rigs, prototypes, or complete vehicles can begin virtually, with the same workloads moving to physical hardware as it becomes available.”

The integration between RemotiveLabs and Google’s Horizon has already been accomplished, according to Sigurdson. The companies are only now announcing it.

But it’s all another step toward using AI, cloud computing and collaboration to boost speed, efficiency and quality into the vehicle development process, pulling together a wondrous technology toolbox, as expressed by SimScale’s David Heiny.

“I believe engineering velocity is the ultimate, if you will, way to be competitive and to be a growing, healthy company.”