One of my favourite IT industry buzz words is fog computing.
Initially coined by Cisco Systems, fog computing also goes by the name ‘distributed computing’ or ‘edge computing’ – never accuse the IT industry of not having enough terms for the same thing.
The idea behind distributed, edge or fog computing is that sometimes cloud computing (another term) is not always suitable for jobs where internet connections are unreliable, sending things to a cloud server wastes valuable milliseconds or lots of unnecessary data is being gathered.
Cisco’s Wim Elfrink discusses fog computing in the interview he did with Decoding the New Economy earlier this year.
For many internet of things applications, edge computing makes sense as it means devices aren’t reliant on being constantly connected and the local computer can filter noise out of the data stream being sent to head office.
Another big advantage is in saving internet costs; rather than having say all the sensors on a car having their own expensive mobile connections they instead report back to a central computer which sends relevant information back to the dealer or manufacturer.
The car is also a good example of needed immediate responses; if the airbags are deployed, the onboard computer system has to make a lot of critical decisions immediately. Waiting for a response from a server on the other side of the world could cost lives.
For computing is not only a good example of how technology vendors make up their own catchy names for concepts, it also shows that there are limits to every solution. In this case, situations where cloud computing is not the answer.