The reason for the kick against the CAN bus


// php echo do_shortcode (‘[responsivevoice_button voice=”US English Male” buttontext=”Listen to Post”]’)?>

There is a glitch in the automotive electronics system architecture. It’s been running for nearly a decade and is now picking up speed. This mostly happens with BEV startups as they have no historical limitations or ingrained favorite designs and can start with a clean sheet of paper for their electronics architectures.

Tesla has shown the benefits of starting with mostly clean electronics design that uses a software-centric and OTA-based system architecture. I think everyone understands the benefits of remote software updates and the OEMs are investing in providing OTA software.

You have to do a lot more and essentially throw away your current playbook via ECU and CAN-based networks. Instead, they have to switch to an electronic system architecture based on Ethernet and a so-called service-oriented architecture (SOA) – the earlier, the better. You could also call this a clean system architecture for automotive electronics. This will take many years, but any new model should use the clean sheet architecture approach.

This column examines this topic and explains why I think it is important that the major auto OEMs get on this system architecture change train. The following table is a summary of the topics to be discussed.

(Click on the image for a larger view)

What is clean sheet architecture?

Clean sheet architecture means using the best new technology available and ignoring older system architectures and their limitations and disadvantages compared to newer technologies. This is always a difficult decision as you will not be leveraging the knowledge base in systems, design, supply chains, testing, maintenance, human resource expertise and other valuable infrastructures.

You will need to develop new infrastructures for the clean sheet architecture to replace the existing support structure, which will be time consuming. Therefore, the clean sheet architecture development will be done primarily for new car model designs. With the move to BEVs and many new designs, this is a good time for OEMs to move to better electronic system architectures.

There is good agreement about the properties that a new automotive system architecture should have. It has to be a software-centric system architecture at all levels that is organized around domain control devices. This means a hierarchy of reusable software platforms connected by application programming interfaces (APIs) – and used in every ECU where possible. The API-connected software platform is synonymous with the SOA.

SOA is widely used in the IT and cloud industries and offers great advantages. It is the basis for Amazon AWS, Azure and almost everyone else that uses cloud platforms. Many of the SOA benefits apply to automotive electronics systems.

Ethernet is the choice for new bus architectures in automotive electronics. There is a wide variety of versions of Ethernet in terms of speed and physical properties, with more to come. These decisions should meet all automotive requirements and usually with multiple benefits from speed and weight to cost – if not now, then in the near future.

A major advantage of Ethernet are better properties for implementing cybersecurity. Because all future car architectures will require integrated cybersecurity and OTA updates, Ethernet is the first choice.

Problems with the CAN bus architecture

The Controller Area Network (CAN) bus has been a workhorse for the automotive industry for about 30 years, but its capabilities are now far surpassed by a family of Ethernet versions. There have been many CAN improvements with CAN FD that offer significant benefits. CAN-FD was released in 2012 and is a good short term solution, but not a long term solution. CAN-FD is not good enough to maintain the usual dominant design wins of CAN in future system architectures.

CAN bus cabling.

The CAN bus speed lags behind Ethernet. FlexRay can be replaced, but the experience base and future improvements cannot compete with the dynamics of Ethernet. Cybersecurity weaknesses have been mentioned and will be an increasing disadvantage.

The main advantage of the CAN bus is the enormous experience base and support structure from part to human. These benefits will wear off over the next few years, only postponing the inevitable.

SOA and Ethernet benefits

SOA defines how software platforms and components communicate with one another using APIs. A major advantage of SOA are reusable and interoperable software platforms via service interfaces. These advantages have created a large SOA ecosystem in the high-tech industry.

The benefits of SOA have spurred growth in chips, systems, and software platforms, and led to rapid expansion in cloud and SaaS platforms. The automotive industry is already seeing the benefits of these SOA-based SaaS and cloud platforms.

AI development also relies on future SOA improvements and technological advances. AI segments such as machine learning and neural networks will increase their influence on the system architecture of automobiles. Autonomous vehicles are particularly dependent on AI and SOA in their high-performance computer systems.

SOA Ethernet innovation potential

The SOA and Ethernet systems receive enormous investments and the resulting innovations from various industries. This is already leading to much more innovation for the automotive industry compared to what happened when many car systems were proprietary to the automotive industry.

Cross-industry investments in SOA Ethernet systems will greatly benefit auto system architectures such as software, SaaS and cloud platforms. Similar advances and innovations are being realized through the computing chips, memory chips, sensors, and other technologies used in auto electronic systems.

OEMs with clean sheet architecture

The companies that use a clean sheet architecture based on SOA and Ethernet are limited to BEV startups. They followed Tesla’s lead, and some surpassed what Tesla did, because they later started with more choices in Ethernet versions and larger ecosystems for SOA.

Autonomous vehicle startups also use SOA Ethernet system architectures – at least for their AV electronics systems. The question is when the traditional OEMs will follow suit and at least carry out test cases if they have to convince of the benefits.


Whether I call it clean sheet system architecture or SOA & Ethernet based system architecture, it’s only a matter of time before it becomes dominant. The benefits of SOA Ethernet are too great, and it makes sense to make the switch as soon as possible.

The major automotive OEMs have had several options to move to the SOA Ethernet system architecture with BEVs as they require major changes to the powertrain system architecture. I think the flywheel momentum effect of traditional designs makes change difficult – especially given the urgency to compete with Tesla and other BEV startups.

Perhaps the OEMs are busy introducing SOA Ethernet architectures into their development programs and we will know in a year or two. In the meantime, the BEV startups are building up system architecture advantages that make them stronger competitors.

The comparison of the SOA Ethernet malfunction in the auto industry with the iPhone malfunction in the smartphone industry may be a bit far. I think it’s in the same class of glitches but it will take longer to play. At least executives at major OEMs understand the implications of the iPhone and will hopefully pay more attention to the potential disruption to the SOA Ethernet architecture and the benefits to the companies that get on this bandwagon.


Leave A Reply