It was never meant to be an exclusive deal, but BMW and Daimler (the parent company of the Mercedes-Benz brand) had planned to work together to develop advanced driver assistance systems, highly automated driving on highways, and automated parking by 2024. However, after a detailed review that included suppliers from their respective supply chains, the two companies have concluded that the cost to develop a shared technology platform would be too complicated and expensive.
In providing this reason for ending the partnership, the companies also noted the current business climate – referring to the massive impacts on the automotive sector due to the Corona pandemic. New car sales have plunged in recent months, with drops in the EU of 52% in May (compared to the previous year) and 78% in April . Just last week, BMW revealed plans to cut 6,000 jobs in Germany.
However, the end of this particular partnership does not mean that the two companies have put all their autonomous vehicle development plans on ice. Both companies have other active autonomous vehicle related projects and partners – BMW is part of a large consortium that includes Intel, Mobileye, and Fiat Chrysler; while Daimler has a partnership with Bosch.
While the reason provided by BMW and Daimler for deciding to end the partnership – that it would cost too much to develop a shared technology platform – may sound cryptic to those outside the automotive manufacturing or software industries, it actually makes a lot of sense. A recent article from McKinsey explains this well: A typical new-generation vehicle likely has a software architecture composed of five or more domains, together comprising hundreds of functional components in the car and in the cloud. These cover everything from infotainment and ADAS (Advanced Driver Assistance Systems) to mapping, telematics, and third-party applications. Typical OEMs constructing this architecture interact with a multitude of software providers to build various capabilities; in the process, they fill their vehicles with a broad set of development languages, operating systems, and software structures. This piecemeal approach is common among industry leaders because no single software platform on the market can meet all cross-system needs. 
The result of this reliance on so many suppliers providing externally built software systems is that OEMs like BMW and Daimler end up with a daunting task of interface control and integration, creating major challenges in development, security, and performance (Volkswagen’s ongoing troubles with its software systems in its new ID3 electric vehicle  is likely a product of such factors). Therefore, committing to developing a shared technology platform adds an additional layer of complexity and integration – not only would each car company need to ensure integration and interoperability between the different suppliers within their own supply chains, they would then need to do this as well with suppliers between their supply chains.
After a little bit of digging, it also appears that the announcement about the split in the partnership between BMW and Daimler came just a few days after a separate development – that Daimler would be teaming up with Nvidia (one of Intel’s top rivals in the computer chip market) to develop a next-generation computing platform that will support everything from over-the-air software updates to automated driving . Developing a shared technology platform would have meant Daimler would not be able to use the technology as quickly as it should be able to now with Nvidia. According to an analyst at Guidehouse, the solution offered by Nvidia appears to be much more powerful than anything coming from Intel/Mobileye which is what BMW is currently committed to.
I find these developments fascinating because it shows just how quickly the automotive landscape is changing, with the traditional OEMs still trying to find their feet in the shifting dunes. This lies in stark contrast to new automakers, like Tesla and Rivian who, as digital-native companies, have had software development at the core of their product design. Also, unlike the traditional OEMs, these new players have a much more vertically integrated business model. I think these factors may be their biggest advantage in the race towards full autonomy.
Written by Bobby Chen, RISE.
3. 2020-01-16 The case for an end-to-end automotive-software platform.