I read your articles with interest and would like to ask a question, how to competent people in the automotive industry – I periodically encounter a CAN bus and in the process of studying I met information about the prospects for switching to CAN FD. Wikipedia talks about moving to FD most auto manufacturers by 2020, however, I have not found any real articles about the application these days, what do you think the practice of using CAN FD on the market now looks like?
Thank you in advance, Alex.
automotive body SW development
- the development cycle of electronic control units is quite small and takes several years, depending on the complexity and integrity in approaching the development process
- The electronic component base that supports CAN FD was released relatively recently and has not yet been fully tested and implemented
- modern data transfer protocols (CAN-matrices, etc.) of modern cars are not yet adapted to the new long CAN FD message format, there are some nuances. Translation of the electronic network architecture of cars takes a lot of time and money and is usually tied to point number 1
- for a number of reasons (both technical and financial), only cars of the middle and higher class, in which the electronic architecture is widely developed, need the CAN FD protocol. What is the point in CAN FD, a high-throughput bus, when there are only 5 control units on board? Lowcost is still a market leader.
- the economic situation in the personal transport market, a decrease in demand due to market overcrowding and a slowdown in the global economy as a whole.
For these reasons, CAN FD is still not massively implemented. But, nevertheless, due to the lack of a more suitable alternative as a data transfer protocol on board automotive vehicles, due to the real need to increase bus throughput, this standard will inevitably be introduced in the next few years (1-3 years) . In addition, other promising protocols were defeated (FlexRay, Most).
From a technical point of view, the protocol also has some questions. The most important thing from my point of view is as follows: automotive network equipment manufacturers are interested in increasing bus throughput. But in fact, the frequency of message transmission during the transition from classic CAN to CAN FD will not change in any way, since practically no changes in the frame structure have occurred. In fact, only the amount of data in one frame has increased. But, given that the network is divided into messages for their intended purpose, an increase in the amount of transmitted data in one message is not so important as it is important to increase the bus throughput (increase the baudrate) in order to increase the number of control units on one bus, that is, the bus capacity. When switching to CAN FD, the bus capacity, of course, will increase, but indirectly. It will already save network segmentation and the use of other interfaces (LIN, Ethernet) in accordance with their application.
I hope I answered your question.