Nowadays, it's common to have several CAN-Buses, which have, depending on their various functions, differing speeds and protocols. The so-called gateways mediate between the CAN-Buses, because they can communicate with the respective networks. In this case, they behave like normal participants. At the moment (2007), there are still (mostly) two ways for the workshop testing device, or other computers, to contact the control devices. In the most favourable case, the device connected to the OBD-socket is accepted as a new participant and depending on it's access-rights, has access to the respective gateways. Should this not be possible, one or more K-cables on the same socket(s), which, via a control device or devices, can obtain access to the network and thus, also to the other participants.

In the meantime, a certain degree of standardization of the information on the CAN-Bus has asserted itself. If e.g., the engine control device sends the latest engine-strain data via the CAN-Bus to other control devices, this is no longer a signal which can only be intepreted by these two devices, it is in fact, now generally valid and can also be correctly read by devices from other manufacturers. This is why, in most cases, the strain is given in Nm.
If the electrical system is switched off via terminal 15 (ignition), all the buses, except for the comfort bus, are also switched off immediatly or after a short while. The data Bus responsible for comfort must remain in stand-by modus, waiting, e.g., for a wireless signal for unlocking or opening the doors, or for the monitoring of the interior. This saves electricity, indeed, still guarantees short response times to cope with certain situations. For the measuring of silent consumers, a waiting period of approx. 30 minutes is assumed, until all the control devices 'have gone to sleep', unless the manufacturer's state otherwise. 12/09