The What, Why And How Of Edge Computing

We all love our automobiles, don't we? That feeling when you grab the keys, get in the driver's seat and hit the gas is beyond compare! Although, that might change with the inevitable arrival of autonomous vehicles, better known as self-driving cars. Think about it - a car that can maneuver itself through dense traffic, winding roads and even the harshest weather conditions without any human input. Say what?!

We know what you're thinking now; can a computer really drive as well as a human? Imagine cruising down the freeway; you are aware of the cars that are changing lanes, those entering via the on-ramp, the rearview mirror, the speed limit, etc. Can a computer cope with all that at once?

Well, turns out it can! Various cutting-edge sensors are embedded within the autonomous vehicle's body that continuously capture data about the vehicle and its surroundings. This data is instantly analyzed, based on which the car knows whether to stop, slow down or turn. Naturally, this process needs to occur in real-time, or the consequences can be fatal.

So, how do self-driving cars manage to process terabytes of data in a flash? The answer is Edge Computing and no points for guessing; that's what this piece is all about! 
 

What Is Edge Computing Anyways?

Buzzwords aside, the closer, the better! That's all you really need to understand Edge Computing. However, at TechDogs, we take pride in getting into the details! So, to fully understand Edge Computing, let's take the case of an autonomous vehicle.

Imagine a scenario where the vehicle needs to stop instantly. Undoubtedly, this process needs to occur in the blink of an eye – just like you would step on the brakes in the same scenario. The consequences will be disastrous if the car needs to wait for the remote server to process the information sent by the sensors, recognize the danger, send back information about what is to be done and then act.

Step in, Edge Computing - a method to process information as close to its source as possible. The main goal of Edge Computing is to bring computation and data storage closer to the user instead of being sent over long routes to a data center and back again. Using edge nodes, the processing is done locally, latency is reduced and the central server isn't burdened with computational tasks.

Now that we understand a bit about Edge Computing, let's see how it emerged.
 

The Evolution Of Edge Computing

 
Although the concept of Edge Computing is new, we can find its origin in the 1990s. In the same decade, Akamai launched its Content Delivery Network (CDN), which introduced server nodes geographically closer to the end-user, allowing cached content such as images and videos to load faster. Then in 2001, peer-to-peer networks were introduced, allowing files to be shared among "peers" on a network without the need for a central server.

However, the breakthrough for Edge Computing came in the form of IoT (Internet of Things). IoT sensors offered the ability to capture data in real-time and promised to expand the "data analytics" horizon for many industries. However, there was no actual way to process this information due to its instantaneous nature. As a result, increasing importance was placed on processing data locally to enable rapid decision-making. Then, Edge Computing showed up.

Today, Edge Computing is ruling the tech landscape. Almost any application that needs real-time data processing or instant data retrieval will use some form of Edge Computing - from autonomous vehicles to smart homes to cloud gaming!

Interesting, right? Let’s take a peek behind the curtain to see how it actually works.
 

So, How Does This Edgy Technology Work?

It's not just us who finds it vexing to wait for an image to download while using mobile data. If you're traveling, it gets downright frustrating since your cellphone reception gets worse the further you are from the cell towers. However, as soon as you get closer to a tower, the reception improves and your downloads become faster. Now you can finally show Uncle Jeff your prom photos!

Before Edge Computing's heyday, data transfer and processing worked on a similar principle as cell phone towers. Instead of cell phone towers, though, we had data servers. They were essentially physical storage units where data (video, music, image, text, etc.) was stored that you could access through the Internet. The further your device was from a data server, the more time it took to download, stream or process this data. Makes sense, right?

Edge Computing solves this problem by placing data servers closer to where the information is generated. Instead of sending data to a server for processing and waiting for it to be sent back, the entire process is handled locally. As a result, latency is reduced and processing power is vastly improved. Does that mean central data servers are redundant? Hardly! At times the computational power needed is more than the local processor can manage and, in such cases, data is sent to the central server.

Thus, the need for data to always be transported to another server environment is eliminated using edge devices. Faster speeds are always a plus, right? Cause who wants the stream to buffer right before the freekick is taken? #damnslowinternet
 

How Edgy Is Edge Computing?

What Edge Computing does is eliminate one of the most critical issues of modern computing - latency. That's what gives it an edge over Cloud Computing (pun intended!). For maximum clarity, let's look at another example.

Imagine you need to secure a research facility (say Area 51, cause we all like conspiracy theories) and decide to set up motion detection cameras. Given the facility's size, you'll probably have hundreds of cameras streaming live to a central cloud server. This constant stream of data takes up a significant portion of the network bandwidth, not to mention the digital space needed to store these recordings. If a camera detects motion, it notifies the cloud server, which retrieves footage from that camera and instructs the motion detection software to evaluate the security risk. This entire process requires too much time, processing power and storage.

Now, let's be smart about it and implement an Edge Computing setup. Each camera becomes an edge device itself; that is, it uses a processor to run the motion detection software and analyze the footage locally. This would considerably reduce the network bandwidth usage and storage issues as only the required footage would be sent to the central server. With this design, more cameras can be added to the network without overloading the central server. Basically, the entire process is done faster and with lesser computing power.

If that made you appreciate the awesomeness of Edge Computing, hold on till we tell you about its benefits. 
 

Why Edge Computing Is Truly Awesome!

 
Edge Computing solves the following problems and also provides the basis for emerging technologies that require instant data processing:
 

• Speed

  The longer it takes to process the data, the less relevant it becomes. Edge Computing offers almost instantaneous processing, much faster than cloud computing. 

• Cost Savings

  Edge Computing lets you retain selective data at your edge locations, reducing the need to transfer data again and again. This helps optimize data transfer and hence costs associated with it.

• Privacy

  By spreading out your data storage across multiple edge nodes, the data is more private because when you transfer less information, there is less of it that can be intercepted. 

• Power Consumption

  Spreading out processing power means - faster and efficient processing at the central server as well as the edge nodes. 

• Reliability

  Since there is no single point of failure, Edge Computing is highly reliable. 

Edge Computing is becoming more inexpensive day by day, allowing more businesses to use this technology and harvest the benefit of faster processing at a lower cost and higher level of reliability. However, the best is yet to come!
 

Where Is Edge Computing Headed?  

We're all standing on the…edge (you knew it was coming) of a future where Edge Computing will be impossible to avoid. Irrespective of the industry you are in— production, energy or transportation —IoT will significantly impact your business. Using real-time data captured by IoT sensors, Edge Computing will manage and analyze all the data at a fantastic speed while reducing network congestion and cost.

In the future, we will witness a major revolution in the transport industry as Edge Computing will combine with the 5G cellular network. This will allow smart vehicles on the road to interact with one another and their environment at impressive speeds, giving us truly self-driving vehicles.

We will also witness a transformation in the gaming industry as game streaming services will construct thousands of edge data centers for scaling multiplayer online games. (So, you won't be able to blame your upside-down K/D ratio on high latency anymore!) 

We'll also see simple changes, such as improved cellular reception, fewer dropped calls and even faster response times from your smart toaster of the future (hey, if it doesn't burn the toast, we don't care how it works!). We'll see changes in the workplace, as Augmented Reality and Virtual Reality will be used for real-time business conferencing and virtually providing on-the-go instructions to warehouse workers.

However, for now, we’re stuck with boring old cloud systems (boo!) although they’re still pretty awesome for what they do!