An Open Platform for Software-Defined Vehicles: Expectations vs. Reality
January 12, 2024
Welcome to the first installment of Scoping Out the Software Defined Vehicle, a series brought to you by the Eclipse Foundation to take an inside look at one of the top trends in the embedded space — software-defined vehicles. In this installment, we spoke with Martin Schleicher, head of software strategy at Continental, about building an open platform for use in SDVs.
There are always a handful of technologies and buzzwords floating around the forefront of the embedded space and beyond — they promise unimaginable innovations, to be The Next Big Thing in x industry. Technologies like artificial intelligence, the Internet of Things, open source, or software-defined vehicles.
But sometimes, these big umbrella terms have a crossroads that makes them more tangible for users and developers — more applicable to real-life customer needs. Take the Eclipse Foundation, for example, who have formed a working group of several key industry players to develop an open technology platform for software-defined vehicles.
The open-source movement, if you will, has been gaining popularity for years. Lots of people believe in the freedom aspect, the community aspect, and it shows. Open source is all over, being used in all kinds of applications and projects, big and small. But what happens when you want to use open-source software in something like a vehicle, where people’s safety is entirely dependent on the vehicle’s systems running properly?
First of all, you have to build the platform. According to Martin Schleicher, head of software strategy at Continental (which is a member of the Eclipse Foundation’s SDV Working Group), even this rarely goes as planned.
The expectation — or perhaps hope would be a better term — is that you have all your decisions regarding the hardware that will be used in your platform hammered out in the early stages of the product development lifecycle, Schleicher says. Then, with that out of the way, you can build on top of that already-established and mature base according to your customers’ needs.
Reality, though, often has other ideas, and this is where the difficulty lies. Not only is it very possible that your platform concept will “break,” but according to Schleicher, it’s also nearly impossible to even build a platform base with technology that would be suitable to the diverse needs and requirements of different customers.
Well, perhaps open source will be a solution to these difficulties, particularly in the software-defined vehicle realm of today’s automotive industry. Schleicher said that Continental has been implementing open source in applications like infotainment for quite some time but does not use any open-source components at all in safety-critical systems at this time.
But, with this working group from the Eclipse Foundation, that could well be on its way to changing. “With the growing complexity of the ecosystem, of automotive electronics and software, technology trends like Linux and POSIX systems or adaptive autos are being introduced more widely with new vehicle architecture,” Schleicher said. “So open source will be used in a much wider area inside the vehicle, not only in infotainment, but also outside the vehicle for the development tools and environments.
The crossroads of open source and software-defined vehicles is definitely a work in progress, but there’s a lot there to be taken advantage of if it’s done properly, such as the benefits that come with connecting to the cloud — another area ripe with open-source ecosystems. Of course, developers must tread carefully, taking care not to sacrifice anything in terms of a platform’s quality or security in the effort to integrate open source.
Stay tuned for more Scoping Out the Software Defined Vehicle content on embeddedcomputing.com.