Automotive Ethernet: A crossroads for the connected car

January 31, 2018 Brandon Lewis

The term “connected car” loosely describes any vehicle with Internet access. It implies an external network connection, and possibly the provisioning of a wireless hotspot that other nearby devices can join.

But in the age of IoT, is that really enough to consider a car “connected”? We now expect all of our other connected devices to be more than mere access points. We expect all of our other connected devices to do something with connectivity.

This is part of the reason Teslas are so compelling. And the Tesla Model S, for example, has an internal Ethernet network.

The expansion of automotive Ethernet

Ethernet first made an appearance in consumer vehicles back in 2008 when BMW introduced it as a diagnostics interface for reprogramming calibration software on engine control modules. Whereas reflashing these modules had previously taken 10 hours over the CAN bus, only 20 minutes was required to upload the same package using 1 GbE technology. The time savings alone covered the cost of integrating the bus into vehicles, and since then Ethernet has invaded other automotive subsystems such as backup cameras and infotainment[1]. For instance, the Model S is equipped with a 100 Mbps full duplex Ethernet network that connects the center console, dashboard and navigation screen, and another unknown device[2].

In fact, Ethernet has been so successful in automotive applications that it has brought the industry to a crossroads. With the advent of deterministic Ethernet standards developed by the IEEE’s 802.1 Time-Sensitive Networking (TSN) task group, the technology can now be implemented in automotive control applications where buses such as CAN, LIN, and FlexRay have traditionally dominated. This evolution also coincides with the need for higher bandwidth data links in automotive vision and active safety applications, which seem to incorporate more high-performance sensors every year.

Today, suppliers like Molex are brining multi-gigabit automotive Ethernet solutions to market that address these use cases. At CES 2018, the company showcased a fully functional in-vehicle network based on a 10 Gbps Ethernet gateway, which also supported CAN, LIN, FlexRay, and other common automotive buses (Figure 1). The customizable gateway leveraged an Aquantia chip for the Ethernet MAC and PHY, while an Ethernet Audio Video Bridging (AVB)/TSN protocol stack from middleware vendor Excelfore guided data across the network.

Figure 1. At CES 2018, Molex demonstrated a fully functional 10 Gbps automotive Ethernet network. Their custom gateway also supports CAN, LIN, FlexRay, MOST, and other automotive protocols.

Beyond just higher bandwidth, a key advantage of Ethernet in the connected car is TCP/IP. Because automotive Ethernet supports TCP/IP, in-vehicle communications over an Ethernet backbone roughly resemble the external communications between a vehicle and the cloud. As a result, automakers can use a common networking technology across their connected car infrastructure, which both reduces design complexity and enables more seamless over-the-air (OTA) software updates to vehicles deployed in the field. This was on display in the Molex demonstration through a feature called Diagnostics-over-IP (DoIP), which discovered all of the vehicle’s IP-enabled devices, established a route to each, and provided a method for reflashing their software (Figure 2).

Figure 2. The Molex Ethernet network and Excelfore TSN stack enable remote vehicle diagnostics over IP.

Vehicle communications crossroads

As stated previously, however, the automotive industry is at a crossroads. In a market that has historically been defined by long technology lifecycles, automotive Ethernet is in its very early days. Before OEMs and Tier 1s commit to large-scale migrations away from legacy buses, Ethernet must prove itself in terms of cost, reliability, security, and a host of other design factors.

Still, the bar for connected vehicles has already been set by the likes of Tesla and others, and automotive Ethernet appears to be the best path forward. At CES I asked the head of connected cars for a major auto manufacturer whether their new concept vehicle had an Ethernet backbone. He paused for a few seconds before saying, “It will.”


References:

1. "Successful connection on the Model S internal Ethernet network." Tesla Motors Club. Accessed January 30, 2018. https://teslamotorsclub.com/tmc/threads/successful-connection-on-the-model-s-internal-ethernet-network.28185/.

2. "The case for ethernet in automotive communications." ACM SIGBED Review. Accessed January 30, 2018. https://dl.acm.org/citation.cfm?id=2095257

About the Author

Brandon Lewis

Brandon Lewis, Editor-in-Chief of Embedded Computing Design, is responsible for guiding the property's content strategy, editorial direction, and engineering community engagement, which includes IoT Design, Automotive Embedded Systems, the Power Page, Industrial AI & Machine Learning, and other publications. As an experienced technical journalist, editor, and reporter with an aptitude for identifying key technologies, products, and market trends in the embedded technology sector, he enjoys covering topics that range from development kits and tools to cyber security and technology business models. Brandon received a BA in English Literature from Arizona State University, where he graduated cum laude. He can be reached by email at brandon.lewis@opensysmedia.com.

Follow on Twitter Follow on Linkedin Visit Website More Content by Brandon Lewis
Previous Article
Four reasons to embrace the connected car

Remember moving from phone to fax, fax to email, or online shopping? Security and value were early adoption...

Next Article
TI releases MCU family for cost-sensitive power control applications
TI releases MCU family for cost-sensitive power control applications

The highly integrated system-on-chip boosts performance in EV/HEV, grid infrastructure, and industrial appl...