A bright future for CAN

The Covid-19 pandemic has changed our world, not just for a short time. Home office opportunities, for example, will not disappear completely with the coronavirus.

In 2021, CiA will offer most of its education and training services online. Of course, there will be again seminars in the CiA office or on customers’ facilities, when it is possible – perhaps end of this year. Online meetings cannot replace personal encounters. This is true in particular for CiA working group and standardization committee meetings. Trust and personal relations cannot be digitalized. We are human beings; we are not machines.

Nevertheless, automation penetrates our lives in many respects and directions. Automated driving is still on the agenda, for example. All these modern automation systems require increasingly embedded networking. This is the business of the CAN community. Already 35 years ago, the Classical CAN protocol was introduced. With the ISO 11898-1:2015 standard, the CAN FD protocol saw the light of day a couple of years ago. And this year, CiA will release the CAN XL specification. CiA also plans CAN XL plugfests to test the interoperability of different implementations.

The increasing demands on embedded networking require not just more bandwidth, but also more design flexibility. This is one the challenges, we are facing. The CAN XL approach is well prepared for this. Separation of frame priority and addressing by means of the acceptance field is just one example. The SDU type field indicating the used next higher OSI (open systems interconnect) layer is another one. With the CAN XL SIC (signal improvement capability) physical media access it is possible to realize bit rates of 10 Mbit/s and more.

Embedded networks are invisible for the end-user. This is intended. But for the marketing of embedded solutions, this is a challenge. Original equipment manufacturers regard the used embedded network as a secret – a very big mystery. Nobody tells us that they use Classical CAN or CAN FD embedded in the product, which they are offering. The same will happen with CAN XL. This is also true for higher-layer protocols based on CAN data link layer technologies.

CAN FD Light is a protocol implementation originally developed for smart lighting systems in road vehicles. It is used deeply embedded in adaptive headlights, for example, linking thousands of LEDs. The carmakers do not care about the implemented deeply embedded network as long as the entire mimic they buy provides an interface for the chosen in-vehicle network technology.

The future for CAN is bright. However, it goes one step deeper into the automation systems. Classical CAN and CAN FD become increasingly a device-internal or sub-system network technology. CAN XL is the option for the system level used as a backbone in automation systems. The good news is that deeply embedded systems in devices and sub-systems are normally high-volume applications. This makes the CAN chipmakers happy.

The world is not black and white – there are many grey shapes. CAN with its three generations of data link layer protocols and its wide range of physical media options, provides sufficient flexibility to meet the different requirements of many embedded and deeply embedded network applications. This ranges from in-vehicle network to sensor networks on the ocean ground and to the embedded systems used in satellites. Design flexibility and scalability is one of the most important features required in embedded networking.

What is missing that are consistent higher-layer protocols for CAN-based networks. Classic CANopen and CAN FD are suitable for many embedded control applications and J1939-based profiles are well established in commercial vehicles. Also, the passenger car industry has developed several CAN-based higher-layer protocols (e.g. the ISO 15675 series of transport layers as well as the ISO 14229-2 CAN on Unified Diagnostics Services profile). What is needed additionally is an approach to run these legacy solutions on the same cable. CAN XL is prepared to do so. Migration paths are needed, too. CiA is committed to support the development of standardized solutions.

To summarize, Classical CAN and CAN FD are not in competition to CAN XL. Of course, CAN XL can be used for embedded control applications as Classical CAN and CAN FD. But it is not optimized for this. It is designed for backbone networks in control systems.