Over a million developers have joined DZone.
{{announcement.body}}
{{announcement.title}}

What Is Edge Computing?

DZone 's Guide to

What Is Edge Computing?

And does it have a promising future?

· IoT Zone ·
Free Resource

When edge computing started making rounds of the Internet, some of the articles came up with statements like:

“The Edge will eat the cloud”

“The End of Cloud Computing Era”

“Edge Computing will blow away cloud”

“Edge will replace the cloud”

However, these headlines might sting a little because you just began to understand the cloud and computing. And now, the digital world seems to have come up with a more advanced solution, i.e. edge computing.

So, are these statements true?

Well, no. Edge computing is just the extended version or advanced functionality of cloud computing.

“The edge computing market size is expected to grow from USD 1.47 Billion in 2017 to USD 6.72 Billion by 2022, at a CAGR of 35.4 percent during the forecast period.”

Why Is There So Much Hype Around Edge Computing?

In this article, we are going to answer the following questions:

  • What is edge computing?
  • What does it bring to the table?
  • How does it benefit the business?
  • Will it really replace cloud computing?

Let's get started.

Image title

What Is Edge Computing?

According to Wikipedia:

"Edge computing is a distributed computing which brings computer data storage closer to the location where it is needed. Computation is largely or completely performed on distributed device nodes."

Let’s simplify it a bit more:

"Instead of holding and processing data in a cloud or central data center, edge computing processes the data at the edge (at which it was originated) of the network."

It makes the system and application more efficient by removing the components and data services from a centralized server and placing the data to the closer to the edge.

This reduces the flow of traffic to the central server and provides real-time local data analysis.

What Are Edge Devices?

Edge devices are any device that produces data. It could be sensors, industrial machines, routers, WANs, switch, etc.

They will act as small data centers that are able to communicate important information with each other.

Why Do We Need Edge Computing?

It is universally true that cloud computing helps to achieve mass centralization, data processing, and enables focus around storage and scaling.

But challenges of latency still exist.

When data travels miles from the cloud data center to the end-user, it goes over thousands of miles to the end-users; so delays are considerable. Thus, devices are unable to take instant action.

A split-second in the delay of data going to the cloud and back to the device can result in a serious problem.

Latency is not completely avoidable.

Understand By Example

A Self-Driven Autonomous Car

Safety is always the first concern of connected and autonomous car. Driverless cars need to ensure that they are keeping to their lanes, recognizing and stopping at the red lights, decelerating by identifying pedestrians and other obstacles and much more. All this requires a massive amount of sensory data in real-time. The car needs to send terabytes of data to the central server and then receives the response and then act.

In case of any mishap, the car needs to take crucial decisions. if we send such massive data to the centralized cloud servers to process and get the response it can result in disaster, as It is not fast enough to respond to the immediate dangers in real-time.

In such a scenario, edge computing can help.

It reduces strain on the network and provides better reliability by reducing the time lag between data processing and transferring the data to the vehicle. It can perform data analytics and speed up the analytics process, allowing the driverless car to take instant action.

“It’s not in a cloud; it’s not in data centers, its right in the computer in the car. Engines can learn how to drive themselves without being reliant on connectivity,”
-Thomas Bittman

Image title

Why Edge Computing Is Important for Businesses?

Businesses can benefit from the reduced latency with the instant access of real-time data in remote locations. The main purpose of edge computing is to decentralize data handling.

If you are collecting the data from the cloud and unable to provide the expected speed and agility, then edge computing is surely the best choice!

To learn more about the advantages of edge computing and it’s real-life use cases, read on.

Advantages of Using Edge Computing

Cost-Effective Data Processing Solution

When you think to develop IoT product, the cost is always the concern due to its network bandwidth, data storage, and computational power. But, edge computing can lower the IT cost by processing and analyzing the important data locally.

When data is stored and processed at the edge, you don’t need more cloud storage. As smaller operations are involved and less data management expenses of the local devices, it will reduce the data transaction costs.

Security at the Highest Level

When the number of devices is connected, it increases the chance of an overall attack on the networks. Traditional cloud computing is centralized so it is vulnerable to DDoS attacks and power outages.

But, edge computing distributes processing and storage across device and data centers, which make it difficult for any single disruption to take down the networks. Moreover, edge computing filters out sensitive information and transfers only non-sensitive data for further processing to adhere to strict security and compliance. As there are fewer data intercepted during transit, it is easier to implement security protocols.

Reliable and Uninterrupted Connection

As microdata centers locally store and process the data, IoT applications consume less bandwidth and work even when the connection to the cloud is affected. Edge computing operates in the conditions with the limited connectivity so business operations can be carried out without worrying about the data loss.

As many edge computing devices and edge data centers are connected to the network, there are least chances for one failure to shut down entire services. Data can be routed through other paths to ensure that users can have access to the product and information as and when they need.

Interoperability Between New and Legacy Devices

With edge computing, it is easy to convert communication protocols used by the legacy device into a language that is easily understood. This makes it easy to connect legacy tools with modern IoT platforms.

Thus, businesses can immediately capture the insights across their operations without investing in new equipment.

Flipping Another Side

Although technology increases productivity, it has its drawbacks — and edge computing is no exception.

  • As this technology is new, many organizations are still learning to utilize the data from the edge locations.
  • Edge computing analyzes and processes the only subset of data, discarding raw information and incomplete insights.
  • It doesn’t verify whether the user was authorized or not
  • Implementing an edge computing strategy can be complex.
  • As several new intelligent devices are connected to a network, configuring all these devices in the right way result in the risk of human error

Use Cases of Edge Computing Deployment

Video Conferencing

Most of us have experienced occasional shortcomings like video delays, bad video quality, frozen screen-share, and much more. Delivering high-quality video is complex due to the slow response back to the cloud.

By placing the server of a voice close to the participants, these kinds of quality problems can be reduced drastically. Edge video server would enable the resilient and responsive user experience for participants.

Smart Cities

In urban areas, sensors are placed for collecting data on traffic patterns and utility usage. It gathers a massive amount of information every day. All this information must be collected, stored, and analyzed before responding to the problems.

Traditional cloud solutions aren’t able to provide immediate response for the device operating on the outskirts of the network. Edge computing makes it possible to respond to the changing condition in near real-time by collecting the data to carry out basic processing task.

Oil Rig in the Ocean

An oil rig has thousands of sensors producing massive amounts of data, and most of them are unimportant yet can help determine if the system is working properly.

It is not necessary to send every bit of data over the network as soon as it is produced. Instead, the edge computing system can compile data and send daily reports to the cloud for long-term storage. By sending important data to the network, the edge computing system reduces the data traversing. It can even help in predictive maintenance to identify the potential breakdowns before they impact the production.

Top Companies Using Edge Computing

Edge computing has not yet become main-stream for business, but there are a number of companies who have already adopted and are streamlining productivity through edge computing.

Amazon

Amazon announced the general availability of AWS Greengrass – the software that powers local edge gateways and appliances. It delivers three benefits:

  • Reduced latency between the devices and data processing layer
  • The drop-off in the bandwidth costs
  • Achieved compliance and security by holding the sensitive data locally

Dropbox

Dropbox, leading file storage and sharing company, has developed its own edge network to deliver better connectivity and faster file access for its customers. The company shifted to edge to provide consistent and reliable service to all its customers.

With this shift, speed was increased by 300% and performance was improved to the next level. With edge computing, download and upload speed were increased throughout Asia and Europe and latency were improved by 5x.

General Electric

General Electric supports a large number of edge devices – up to 200,000 connected devices, which enable faster and more efficient processing at low latency.

Edge solutions drive real-time insights of major industrial assets while also connected securely to the cloud. This technology delivered the outcome that resulted in increased revenue, efficiency and reliability across multiple industries.

What Do You Mean, It's the End of the Cloud Era?

Edge computing doesn’t mean the death of cloud computing.

Edge computing is more of a complement to cloud computing; it can’t replace cloud computing because there will always be the need for central processing. It needs a cloud to configure, deploy, and manage IoT devices for analyzing large datasets from different data sources.

It makes a sense to perform IoT analysis at the edge when immediate action is needed as well as faster data processing, but the data will likely need to be collected and centrally stored.

Ultimately, cloud computing will remain an integral part of any edge computing environment.

Edge Computing: A Promising Future?

Edge computing is in its infancy stage, and it is difficult to make any real statement on what the future may hold. However, from the above highlights on edge computing, it seems that edge computing is becoming an integral part of developing IoT applications.

Edge computing is much more than simply redistributing data to endpoints. There is no denying that it will enable businesses to perform advanced analytics in new, more profound ways. The rise in localized data centers may also make it easier for organizations to expand their network reach.

Going forward, we may start to see more and more applications in healthcare, AR, VR, drones, smart cities, and other sectors. So as far as I can tell, edge computing it here for the long haul — it's here to stay.

What are your thoughts? Please share your feedback in the comment section below.

Topics:
edge computing ,edge computing applications ,iot ,edge ,data ,data processing ,smart cities ,iot applications

Published at DZone with permission of

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}