Current conventional devices and sensors are designed to collect data from machines and other equipment and back-haul this to the central cloud or a datacenter where the data is analysed and the results sent back to the machines/equipment, based on which a decision is taken.
The deal with such an arrangement is that it takes a bit of time to receive the results. The delay can be anywhere between a few seconds and a few minutes and also depends on where the central datacenter/cloud is physically located. This also means that every bit of data collected has to be sent back to the cloud – putting a strain on storage resources. And all the data gathered must be filtered, sorted and analysed, thus requiring a lot of processing power.
There are many situations where such delays can’t be afforded – mission critical projects, life threatening situations, disasters, accidents, etc. In such cases, split-second decision-making becomes crucial. And this is what edge computing enables.
Edge computing is a development that allows for devices/sensors, etc. to be designed with storage and processing capabilities. Data is processed and the results generated at the source of the data itself. As a result, there is practically no latency, the amount of data being transported is lower and the storage and processing power at the datacenter/cloud is optimised.
Connectivity is a fundamental requirement for this form of architecture. Any disruptions to the connectivity layer would mean a complete breakdown of the system.
In that sense, edge devices can be considered as upgraded and a more intelligent version of IoT devices. And for this reason also, you will notice that the application areas of edge are not too different from that of IoT.
Cloud technologies and edge will complement each other – the former will be leveraged for long term storage and historical analysis; the latter, for real-time processing and faster decision-making.
Is Edge the device or the location?
It is important to understand that the device is located at the edge; it is not the edge itself. This is because the edge means different things to different people. For a network player, the cell tower or a small cell would be the edge; for a cloud service provider, it would be the datacentres or micro-datacenters. This is called the Infrastructure Edge. For a retail consumer, the edge would be the range of devices they own – smart phones, smart appliances, autonomous cars, etc. For companies, the edge would refer to their shop floor, supply chain, etc. This is called the Device Edge.
Edge computing ecosystem is characterised by a diversity of devices, in a distributed environment (spread over a large geographic area), enables distributed computing, real-time decision-making and is autonomous – that is, can work independent of the core network or internet connectivity.
To know more, please read our report: “Edge Computing: Towards a More Distributed Future”, Nov 2019