The Imitation Game: Do it the Silicon Valley way?

Last week, I had the chance to participate at the #Intersect conference at Mountain View, hosted by Applied Intuition at the Computer Museum, located directly at the 谷歌 campus. On the quest for the "SW defined" vehicle, the US startup founded in 2017 by Peter Ludwig and Qasar Younis is one of many candidates who could contribute to a scaling SW stack. The huge interest of OEMs into this kind of startups has one reason only: Their own failure in getting the SW stack done so far.

The ingredients? Small, but capable teams. Experience in pure SW development. High speed in development and showcasing. And tons of money: Applied Intuition with only 500+ employees just got 550m$ in its last investment rounds, and has an evaluation of 6b$.

Compare this to ZF, Continental, Valeo and Co: Market caps from 4-10b$, ca. 20.000 employees in R&D, and development budgets around 2-3b$ each.

So why are investors looking so differently to these two types of companies, and have been the investments of OEMs and 1st tiers on becoming "agile", "digital" and "SW defined" wasted?

Misunderstanding #1: Agile

It's funny to see that almost all carmakers have invested heavily into SAFe (less on LESS) frameworks in the past decade - because agile is better.

Fun fact: No, Tesla is not working "agile".

There is a fundamental misunderstanding and bias. Yes - the new kids on the block are working extremely fast and iterate like hell. They work in cross-domain teams and show great flexibility. And they listen to customer feedback carefully and directly. They are following very clear principles, and they are pretty different from an ASPICE process catalogue.

But OEMs much to often are not copying those first principles and paradigms - they are looking for an alternative framework to their current complex governance, organization and process landscape.

SAFe is not agile per se - it's a tool to structure big organizations, but it is never a surrogate for the basic principles. Most of the new players are not eager on installing structured frameworks, invest into trainings and so on - they just work after agile principles.

Don't get me wrong: those principles are the foundation of any SAFe training. But the structured approach of SAFe is sometimes far to similar to established OEM structures, and when the former team lead is now the scum master of the same team, you know you're into trouble.

Misunderstanding #2: Digital

The digital ecosystem - e-commerce, data harvesting, subscription models, etc. - has been like an attractive honeypot - but for both sides.

The carmakers were peering at the wonders of digital profit pools, the market caps of the famous 7 reaching heights never seen before. Their way to really scale things (investing once, but earning continuously)

On the other side, the so called hyperscalers are in dire need of the last remaining resources: customer time. 23 hrs of the day are already captured: Customers (or better: data suppliers) are monitored by their laptop, smartphone or tablet. Or even by their wearables at night time. But that one hour of driving around in their cars was missing.

Both views are wrong, as you know. But how then can Applied Intuition reach an evaluation of 6b$, just with the promise of an SDV stack and ecosystem?

Misunderstanding #3: Product

What is the difference between Ariane 6 from European Space Agency - ESA vs. rockets of SpaceX or Isar Aerospace? From a technical perspective, a lot. Also from a total cost of ownership (TCO) perspective. But the root cause is some very different mindset: The insight that it's not about creating the perfect product, but the perfect product creation system.

It's not about the perfect product - it's about a perfect product creation system.

If you create the perfect product, you can deliver a great rocket. But only each 6 months, and after 15 years of development.

If you have a perfect product creation system, you will be capable to deliver a product each 2 weeks as Isar Aerospace is planning. Or refuel a rocket and relaunch after 2 hours, which SpaceX aims for their Starship.

The same is true for Automotive Software.

If you try to deliver a perfect product after 36 months of development, well - you could miss the market. If you are able to develop, integrate and deliver a new feature each day, the product quality of today is not so important - you will have plenty of of time to improve, until your competitor will deliver.

It's important to emphasize that here we are talking about all aspects:

• The product architecture itself, allowing flexibility
• The processes and tool chain of R&D
• The subsequent processes in manufacturing, sales, after sales, repair etc.
• But also the supporting processes such as purchase and HR.

The weakest member of a chain can make it break. If deliver SW like hell, but your purchase department needs 6 months to onboard a new supplier, it's wasted. If you want to refuel a rocket within 2 hours, you need to take care of your pump.

Promising "shift left" - delivering "shift right"

The current vehicle architectures are not a failure, to be precise. The current cars from Mercedes, BMW & Co. are great in many aspects. But most OEMs still struggle to get cost down, and speed up, and manage the complexity of a product evolving over lifecycle.

The fall-out can be seen in many vehicle launches. Delays of more than 2 years, missing key features, and lack of maturity in the field. On the other side: super-fast market entries of new competitors, but lack of maturity, sales and after sales structures.

The rise of Rivian, U Power, Applied Intuition and their kind

In this challenging setup, OEMs are looking for ways out (or at least a plan B to secure their own developments). Here are a few examples:

• 大众 declared to invest 5b$ into Rivian and a Joint Venture with Rivian, aiming to take the Rivian E/E stack for its vehicle fleet (excluding battery/powertrain and autonomous driving for the moment).
• U POWER Tech claims to deliver its electric "skateboards" (the chassis with powertrain and battery) to 50 out of 150 Chinese EV OEMs, delivering a modular and standardized system (yes, there are plenty of them in China!)
• Applied Intuition talks about 18 OEMs as customer of its stack (embedded SW plus development kit).
• And much more existing or emerging platform approaches, including also many open source projects.

These approaches are not for buying just a technical platform, but also to benefit from an ecosystem of development environments, tool chains and automations, and also to get access to leading SW experts and architects. They also aim to share the burden of development cost and resources, as well as the future maintenance of the stack.

Circling back to the Intersect 24 conference: It was a blast.

At most established Automotive conferences you will meet a typical mix of OEMs, 1st tiers, engineering service providers (and many consultants, to be honest). Plus some startup founders for proof of innovation.

At Mountain View, the OEM speakers were flanked by OpenAI founder Sam Altman and management members, by venture and seed capital firms, by GenAI scientists with deep technical presentations, and by robotics legend Sebastian Thrun .

Don't imitate - merge best of two worlds

For most OEMs there will be no way of re-inventing the wheel (aka the SW stack) by themselves. Cooperations and sharing of tech stacks will be essential to refocus on what matters: Customer experience, reasonable cost (and pricing!), value creation which fits to realistic capabilities and key strengths.

There is a "but". If OEMs are treating their new partners as they handle their suppliers, they will fail epically. And if the new players ignore the complex realities of Automotive business at scale, they will do the same.

So, let's learn from each other. But let's not copy just some agile kitsch or "smartphone on wheels" nonsense. As always, reach out to me or my colleagues to exchange on how we can join forces on this crucial journey.

Juergen Reers Philipp Kupferschmidt Dennis Rothhaas Christian Levels Sebastian Angerer Modar Horani Stephen (Yizhou) Xu Amarnath Bharadwaj Markus Muessig Wolfgang K?cher Bettina Blum Malte Becker Sven J?bges Fabian Schuette Dominic Craciunescu Liam Friel Raffaele Menolascino Hans Loes

Yvonne Bernerth