The recent announcement of multiple new members and seven new projects under the
Software Defined Vehicle (SDV) Working Group is a major milestone for the working group,
which was established just a few months ago. And it’s a great validation of our code-first
approach to automotive software development.
Traditionally, established players in the automotive industry have taken a standards-driven,
hardware-first approach to development. Our approach is different. Rather than creating
specifications and architectures for an open source in-vehicle application runtime stack, we’re
starting with code.
To that end, we ask our members to contribute significant source code artifacts to the initiative.
There’s been a strong response to our strategy: membership has grown by 30 percent over the
last few months, with Cariad, Eteration, FutureWei, NXP, and Toyota all becoming members of
the working group. We had our first code contribution day on June 30, which kicked off our first
The first step for the working group will be to see how we can combine the new projects. For
the members, contribution day was a chance to share information about their projects, and an
opportunity to meet one another. We’re hoping these exchanges will be the foundation for
strong collaboration over the next months and years.
Software Now Leading the Hardware
The changes we’re seeing right now in the automotive industry represent a paradigm shift.
Over the last 30 years, car manufacturers have generally developed hardware first, so the
software had to follow the hardware. Because the automotive industry has always been driven
by cost, the hardware never had the performance to let you add features later.
Now, we’re shifting to a new model that’s much like the mobile industry, where the hardware
follows the software. The idea is the future of the automotive industry won’t be defined by
pieces of hardware like speakers in vehicles. Instead, speakers will be a means to an end for
software features, which will likely be sold to customers as part of a monthly subscription.
Emergent, Not Industry-Defined, Standards
In an industry that used to be driven by several traditional players, we’re seeing new entrants
who have a totally different perspective, especially on software. Their perspective aligns with
modern technology development.Many modern technologies, including AI and cloud technologies, aren’t being developed by
following the old-school process of going from a standard to an implementation.
Official standards defined by standard-defining organizations typically take two to three years
to develop. However, if a given platform is big enough and there are enough adopters, it
becomes a kind of de-facto standard.
For competitive reasons, this kind of de-facto standard creation is only possible for the non-
differentiating parts of a software stack. But if there are enough non-differentiating aspects to
the software, you can build a community around the platform. That’s what we’re hoping to do,
not only with the in-vehicle application runtime stack but with automotive tooling software as
An Iterative Approach Accelerates Innovation
We’re also aiming to have very fast time to market. We don’t want to spend two years
developing a specification. We want to take an iterative approach where we develop something
and see how it works. If it works, we keep it. If it doesn’t, we try something new. This approach
is closer to the agile development that’s already used in the automotive industry, but usually
ends once the software is in the car. We want to take that agile process beyond that point.
Our goal is to deliver usable open source code for in-vehicle software across vehicle models,
product lines, brands, organizations, and time. This will increase the speed of innovation and
production and the ability to scale the production of software-centric vehicles while making
new vehicles easier and more efficient to design.