Automotive News Europe — 2024-01-31
Automotive Industry
Automakers creating software-defined vehicles (SDVs) need to decide which control points they want to own in the technology stack.
Deciding factors include creating a user experience that aligns with their business model, as well as how much they can afford in the long term.
Developing in-house technology stacks requires substantial investment.
The development of a car operating system can cost an automaker about $4 bn while lifetime maintenance of the system costs $6 bn, consultancy Accenture said in a recent report on SDVs.
“Looking at potential sales volumes, only a few automakers appear to have the financial backing to establish a full-stack approach themselves,” the report said.
Automakers also need to decide which capabilities they can provide. For example, Tesla is pursuing the full-stack approach, owning and controlling more layers than previous automakers.
A full-stack approach offers automakers the potential to generate the highest revenue, win customer loyalty, and build and own the entire product, from hardware to the cloud back-end.
Nevertheless, this solution necessitates major investment, carries the greatest risk and can be complex, with one key challenge being how to operate a complicated ecosystem over the vehicle's life cycle.
Alternative approach: Work with Google, Amazon
Other automakers, such as Mercedes-Benz and BMW, are opting for an open-source approach, which provides automakers with control over most layers, while they partner with external providers, such as Google, Amazon and Microsoft for the remainder.
Google's bundling of different services has maintained that company's ownership of a significant chunk of the market. SBD Automotive forecasts that nearly 20% of new cars sold in the US in 2032 will have the suite of services known as Google built-in: Google Maps, Google Voice Assistant and the Google Play store.
Polestar, Volvo, General Motors, Renault and Honda sell vehicles with Google built-in. Google Maps is a consumer favorite, and the bundling of all the three services provides a tempting incentive to go with Google Voice assistant as well. Amazon is also working closely with several automakers to offer voice assistance. At the 2024 CES in Las Vegas, BMW and Amazon showcased a voice assistant that is based on Alexa's artificial intelligence model, which processes human language at a high level and generates sophisticated responses.
Forming partnerships saves companies from building their own operating system or cloud framework. On the other hand, they still need to invest in supporting infrastructure and in the capabilities necessary to operate specific layers.
Meanwhile, the domain stacks approach features highly specialized services that can also work using hardware and near-hardware software from a third-party manufacturer. This model offers access to end-consumer data but limits control over the quality and experience provided by the hardware.
Finally, the white label archetype provides a platform for others to build on, using hardware, software or a combination of both. This requires new capabilities as well as substantial financial resources in the long term.