EDGE COMPUTING

Bringing “Intelligence” closer to you

All of the things that we think of separately today , comes together in the future . With the number of devices only growing multi-folds and IoT bringing them closer, with internet connection to every other device you touch, feel or use there is going to be massive explosion in data which each of these billions of devices would generate. 

Imagine things like your fridge , dishwasher, washing machine, coffee maker etc having their own internet connection and generating data.

EDGE COMPUTING will allow these things to process the data locally together with the help of AI & ML so that this data makes sense and these things will no longer be just things, they will become important decision makers.

EDGE COMPUTING means in simple words pushing “intelligence” to the “edge” of the network, to the devices which actually gather the data from the beginning. So basically it’s like we want to process the data at the very machines that gather this data from the beginning and make local decisions at these machines itself instead of centrally in the cloud.

In the future various things will be connected to network. With things connected and coordinated together, and coordinated with information it’s expected that even more value will be created. 

All this information connected to each other is referred to as Internet of Things (IoT) and the huge amount of varied information and data generated by IoT need to be processed and responded to in a very short time. 

As we begin to ponder over the idea of bringing the processing and computing to the edge, we start thinking, then what about the whole idea of not having data in your devices and having it in the cloud. The answer to this is , the cloud is centrally deployed on a global level and needs to process an enormous amount of data which can be complicated and/or would require more computing power etc. In addition, as the physical distance between the user/device and network increases, transmission latency increases with it. This results in slower response ultimately stressing out the user. Many a times this response also depends upon the processing capacity of the users’s device.

As the amount of data created due to all these connected devises, is vastly increasing sending the data to the cloud which is centrally located, through the network will create bandwidth issues by burdening the network with more and more data. The 4G network which we have today will eventually not be able to handle the increasing amount of data, thus causing latency issues. Hence edge computing introduces the concept of edge devices which help in doing the computational work and analytics at the device itself and maybe send the summary to central server by the end of the day or in any manner it is programmed to do so. Machine learning comes in play in these situations which help reduce the load on the network.

Let’s take a simple example of surveillance cameras for understanding purpose. Imagine these cameras are working 24 hours , 7 days a week and keeps sending all the data to the cloud/central server, which then records it and stores it. Most of the data is of no use, say 90% of it is just junk, but what if these surveillance cameras itself were programmed in such a way that they capture data when there is a movement. Maybe the cameras can be fed with biometrics technology which can recognise faces. So, if there is an accident and the person tries to escape , his face can be recognised and sent to all the stations. 

The above example may not be perfect fit in all situations, say for example if there is a storm the movements happening will be recorded and that is junk too. The idea is to just explain as to how, intelligence can be pushed to the edge thus saving time and resources.

The aim of edge computing is to move computation away from the data centres or the central cloud and making the devices and things at the edge more intelligent, such as smart phones, smart homes , automobiles etc. 75% of enterprise-generated data will be created and processed at the edge by 2025, up from 10% in 2019* 

Edge computing promises to bring storage and computing capabilities closer to the mobile users , leveraging existing devices to reduce latencies and core network utilisation.

-Lahari Gowda