In a standard wiring scheme, any input needed to have a dedicated transmission line so that it may communicate with a control unit to affect some sort of an output. However, since a CAN-bus system uses only two wires for all of its communication, how then can it affect any kind of output?
As one can see below, the actual process is quite complex. The format of message that is sent from an input is sent with a base format and extended format (each message carries an identifier, whether it’s a dominant or recessive message, data field in bytes and other coding identifiers). Since all of the control units (or nodes) are tied together with two wires, they must receive all of the traffic from all of the other nodes and decide for itself if the message pertains to them. Any node is free to communicate on the bus, but once a priority message is sent all other traffic is stopped and ignored, e.g. in the event of a crash, the airbags need to be deployed; the fuel needs to be shut off; doors need to be unlocked and a help signal can be sent to an offsite monitoring facility (OnStar®).
This system is extremely fault tolerant. During every message transmission, a portion of that message contains a 15 bit Cyclic Redundancy Check (CRC) code. Each node controller maintains a watchful eye over the bus line. And, if the transmission speed is kept below 125kbps, the bus will still function if one of the two wires are severed, e.g. a car crash.
And, if a node happens to go faulty, then whole system doesn't shut down. Instead, the network is built with safety feature that if a node fails, then that node is excluded from sending any further traffic by means of other nodes not "picking" up the traffic.
