The term “cloud computing” was (supposedly) first used to describe the appearance of networked-servers on schematics. These diagrams used circles to represent the outline of a server (essentially the server’s reach within a network). When overlapped, the circles came together to form the appearance of a cloud. But clouds don’t have a perfectly defined edge, just like in reality.
In the same way that cloud computing has abstracted resources like compute, storage and bandwidth, fog computing has become the go-to name to describe the cloud’s final stop: the Internet of Things (IoT), or all the physical devices and controllers that now have the built-in ability to ‘think’ and ‘talk’ to one another.
Fog computing exists as the farthest reaches of cloud computing, the outer edge of a network – the Oort Cloud of cloud computing. Also dubbed Edge Computing or fogging, the technology facilitates the operation of compute, storage and networking services between data centers but also end devices, helping bring IoT (the Internet of Things) into the picture.
As CeBIT puts it, “…in the case where information is needed only locally and for just a short period of time [data] needs to be processed quickly, preferably onsite, without any need to store the data,” reducing latency and the risk of broken connections. If driverless cars are to be a reality, fog computing will be vital to ensure fast and seamless communication of data. In fact, self-driving cars can be considered the best example of what makes fog computing relevant.
Say an automated vehicle slams on its brakes on the freeway. Other nearby vehicles need to be informed of this within milliseconds to prevent collisions. The only way to achieve this is if the necessary data processing and communication takes place using local computing power, or “the fog.” Communicating the data back to base and waiting for acknowledgement from the cloud would be slow and life threatening. Simply put, fogging created itself out of the the IoT’s necessity to have local decision making.
Regulators are already working on establishing clear directives for this layer with the end goal of improving response times, reducing broadband requirements, and diminishing the required storage capacity at data centers.