Cloud Computing had been a buzz word for quite some time. However, certain latency and network connectivity aspects of cloud computing brought a need in processing the data on-prem or near to the premises where the data is captured. This gave birth to a new aspect called “Edge Computing”. While Edge Computing is not a replacement for Cloud Computing, it has certain advantage over Cloud Computing in certain uses cases, especially from a networking, data processing and storage perspectives. Edge computing is already familiar to us in many ways, from your wearable watches to drone-enabled crop management.

What is Edge Computing?
In simple words, edge computing is the practice of capturing, processing and analyzing data near where it is created, i.e., data analysis takes place on a device in real-time. Thus, this computing brings data and compute closest to the point of interaction and instead of journey to the cloud data center for processing which takes more time, more bandwidth and costs more – the data processing happens on the user’s device at the edge of the network.

Where is the ‘Edge’?
‘Edge’ is relative to a topology. A customer device’s edge is not the service provider’s edge. Since the processing of data over the WAN, even it is to a private data center cannot be called as Edge Computing, the edge should ideally be before the data crosses any WAN.
The following picture illustrates the typical devices in an Industrial IoT solution and the ‘EDGE’ is seen before the WAN in this topology :

Examples of Edge Computing
EDGE Computing offers a variety of unique selling points for smart IoT applications and use cases across multiple industries. Some most popular use cases that depends on Edge computing for improved performance, productivity and security includes
1) Voice Assistant technologies like Amazon Echo, Apple Siri, Google Home etc. which requires computational power and data transmission speed for improved performance. There are aspects where customer’s privacy and data security need to be maintained and thus vendors in this technology are enhancing their AI capabilities and deploying the technology closer to the edge so that data is not moving across the network.
2) Autonomous Vehicles are another real-time processing required devices, where without drivers, the vehicles should be capable of reacting to road incidents on real-time. Latency in data transmission between vehicle sensors and cloud data centers can have huge impact on the responsiveness of self-driven vehicles.
3) Predictive Maintenance in manufacturing industries is another strong use case for edge computing as it demands continuous performance and uptime of automated machines. With edge computing, IoT sensors can monitor health of the machines and identifies the time-sensitive maintenance issues in real-time.

Future of Edge Computing

Several researches and developments are happening in AI and 5G connectivity technologies and with the rising demand of smart industrial applications, edge computing is getting matured faster. Edge computing in telecom, referred to as ‘Mobile Edge Computing’ or ‘Multi- Access Edge Computing’ will be a revolution in the next generation cellular network called 5G. Edge computing will be an enabler for broader use-cases in the future with increasing interests in Augmented Reality/Virtual Reality, upcoming 5G radio networks, smart manufacturing etc.

References:
https://alln-extcloud-storage.cisco.com/blogs/1/2020/02/Edge-Compute-Inline-Image.png
https://stlpartners.com/edge-computing/what-is-edge-computing/

Ms. Anupreetha Rugmini 
Robotic Process Automation Solutions and Platform team, 
Technical Product Manager
EY, Trivandrum