Ethernet - the answer when CAN can’t manage on its own?
Article published in the iVT Magazine.
Several technologies from sectors such as the automotive and factory automation industries have influenced industrial vehicle OEMs over time and, as a result, new solutions have been developed that provide collaboration and savings through economies of scale. One of the common denominators where synergy benefits have not yet been exploited in full is the communication bus.
CAN is proven, mature, and widely supported. However, it also has a few weaknesses, such as the limits in its bus and branch line length and its general throughput. Also, if a single bus, be it wired or wireless, could serve all vehicle systems, from control systems to infotainment, then it would simplify and reduce the cost of vehicle construction.
On this basis, and keeping in mind that overall data volumes and real-time responsiveness requirements are likely to grow, the bus performance issue clearly needs to be addressed. At the same time, the benefits of CAN should not be forgotten.
Ethernet, when added with real-time services, can fulfil the requirements of a deterministic, flexible, high-performance, open, general-purpose next-generation bus for vehicles. With throughput of 1Gb/s available today, it can outscore FlexRay, which has quite a limited application base in the high-end car industry. FlexRay also has some critical limits in bus topology and length, and it is not clear how open it would really be beyond the automotive sector. Ethernet is also gaining ground in the research plans of key FlexRay deployers, giving even more opportunity and leverage for a cross-sector technology push.
Ethernet emerged in factories some 10 years ago, with initial concerns on safety, speed, and timing. The issues were tackled with controlled isolation of process network from business network, and by improved network design. Today, it is used widely in automation.
Read the whole article - Carrying the CAN