RemotiveLabs, the Malmö-based startup building the tools that let carmakers develop software the way modern tech companies do, announced Google Cloud is working with them to bring their “vehicle as code” technology to Horizon, Google Cloud’s open-source platform for building and testing Android Automotive OS software.
Horizon is designed to help automotive manufacturers and suppliers build, test, and deploy software for Android Automotive Operating System (AAOS) more efficiently.
Founded in 2020 in Malmö, RemotiveLabs lets engineers treat a vehicle’s systems as code: version-controlled, testable and shareable, closing the gap between how software is built in most industries and how it’s still built in cars. Its flagship solution, RemotiveTopology, is already used by manufacturers, tech companies and suppliers including Volvo Cars, Ford, Horse Powertrain, HERE Technologies – and supported by Arm to enable end-to-end virtual integration.
“With RemotiveLabs, Horizon extends beyond the vehicle’s cockpit and main compute nodes. It enables developers and agents to test the full vehicle in virtual, long before a physical car exists. That foundation helps software-defined vehicle programs move faster. We’re glad to bring this level of completeness to developers building on Horizon.”
said Florian Haubner, Industry Architect Lead, Google Cloud.
Google Cloud’s software-defined vehicle work has so far centered mainly on the cockpit, where Android Automotive OS lets developers build and test infotainment software in the cloud. Horizon now extends that reach to the vehicle’s other main compute nodes. Until now, there has been no feasible way to represent the rest of the vehicle well enough to test against.
“Our goal from day one has been to let vehicle manufacturers own their software, on their own terms, enabling them to left-shift integration testing through virtualization. Working with Google Cloud on Horizon sends a strong signal to the industry that it is time to move away from closed, hardware-centric and vendor-locked development. We’re thrilled to contribute with the vehicle harness and infrastructure needed to make both humans and AI agents productive in software-defined vehicle development,”
said Per Sigurdson, CEO and co-founder of RemotiveLabs.
RemotiveLabs provides an executable model of these remaining systems, the vehicle’s virtualized electrical and electronic (E/E) architecture, letting any node be tested in the context of the full vehicle. This enables manufacturers, Tier 1 suppliers and other technology companies to collaborate around a shared source of truth. Google Cloud calls the combination a “system of systems”:
RemotiveLabs’ virtualization platform, RemotiveTopology, makes that possible by running the same network communication as the finished vehicle, down to how data packets are encoded and addressed, which also lets virtual systems connect to real hardware. RemotiveLabs also brings the two things AI agents need for this kind of development: the vehicle as code in standard version control, and a virtual environment they can control and test in. Manufacturers and partners can therefore bring the AI tools and agents of their choice to iterate and verify progress at system level, just like any other software project.
