Edge computing and IoT are a perfect match for several reasons. No wonder that edge computing is in virtually all IoT 2018 trend reports, just as it was in 2017 as you can read in our article on Internet of Things 2017 trends.

As a matter of fact, many of the evolutions and trends in that article still go for 2018 and the years after that, how else could it be? Numbers and predictions do change but all in all key evolutions aren’t limited to a year of course. However, we do have some updates. As a reminder: in the previously mentioned April 2017 trends and evolutions article we didn’t talk about edge computing but about fog computing and the time has come to explain that

We often have used the term edge computing and, as most people do, used it interchangeably with fog computing. Still, there is a difference between fog computing and edge computing and the exact (technological) ways in which they play a role in the IoT ecosystem, mainly in industrial IoT that is (but not just there). However, correct is correct and, although both terms are about moving intelligence In our article on fog computing we also mentioned that edge computing is not new, whereas at the time of writing about it, fog computing was definitely ‘new’. Some called it Cisco’s marketing take on edge computing in IoT but there is more. In the same article we also said that fog computing is a form of edge computing, in the ‘old sense’ that is. Given the rising importance of edge computing ‘in the IoT sense’ as shows in so much research, time has come to elaborate a bit more on edge computing, fogging and why the heck it all matters. Before we start: do note that edge computing nor fog computing are just about IoT though However, in this article we mainly look at it from that perspective of course. For a more formal description of the difference between both: fog computing versus edge computing.

Edge computing and fog computing: same drivers

Both edge computing and fog computing are strongly on the rise for the same exact reasons: an IoT data deluge.

This IoT data deluge, among others, takes place in the converging worlds of IT and OT (again predominantly Industrial IoT) and occurs in general as we keep adding more IoT devices in the scope of mainly large-scale IoT projects, the industrial markets of Industry 4.0 and IoT use cases and applications where a lot of data needs to be analyzed and leveraged, often also in an IT and OT environment as we, for instance, find them in IoT in manufacturing.

Another example: smart buildings and building management systems where we increasingly look at the building in a holistic and integrated way instead of from a rather siloed perspective of various areas ranging from energy management and power management to HVAC, light control, energy efficiency and much more.

That same holistic view whereby we want to know what happens in buildings as a whole as happens in facility management takes place in other environments. Industry 4.0, Logistics 4.0 and so forth, for instance, are about an end-to-end view on the product life cycle and the end-to-end value chain and supply chains.

The why of moving intelligence to the edge of IoT

 
If you get a lot of data as is the case when you leverage IoT in such end-to-end ways or even in specific highly sensor-intensive and thus data-intensive environments whereby data is generated at the edge which by definition happens in IoT as your data sensing and gathering devices ARE at the edge (think about all the sensors and data they generate in a large oil and gas project where you can have hundreds of thousands of sensor data points across myriad wells but also about all the IoT data in a smart city or large critical power building such as an airport), you inevitable encounter challenges on levels such as bandwidth, network latency, speed overall and so forth where fog and edge computing play a role. In IoT applications with a mission-critical and/or remote component the need for speed and for different approaches such as edge computing is even more important.

Depending on the context and scope of the project you want the data you need fast. Or better: you need the aggregated and analyzed data, in the shape of actionable intelligence, enabling you to take actions and decisions, fast, whether these decisions are human or not. So, you don’t need all that data to store it and analyze it in the cloud but you only want that bit of data traveling across your networks.

You can imagine hundreds of scenarios where speed and fast data is key, from asset management, critical power issues, process optimization, predictive analytics to the real-time needs of supply chain management in a hyper-connected world, the list is endless.

You can also imagine that the more your building, business ecosystem and whatnot thrives on fast data and real-time holistic management in any broader context, the more valuable that data can become when properly leveraged and rapidly analyzed. We do live in times where having the right insights fast enough can have enormous consequences.

Speed of data and analysis is essential in many industrial IoT applications but is also a key element of industrial transformation and all the other areas where we move towards autonomous and semi-autonomous decisions made by systems, actuators and various controls.

That degree of autonomy is even at the very core of many of the desired outcomes and of the goals in, for instance, Industry 4.0 as we move towards the next stage of the third platform which is all about autonomy.

Edge computing and IoT in 2018 and beyond

With real-time information even being a proven competitive differentiator it is clear the in the increasing unstructured data deluge of which the IoT and sensor data deluge is part, traditional approaches don’t fit anymore as we’ll see.

There are even applications and industries where, just on the level of sending data, traditional networks don’t suffice, let alone can be used, for instance because of their remoteness and the costs it takes to send all this data through, for instance, satellite communications.

So, for a mix of reasons (bandwidth, costs, speed, automation, maintenance, predictive analytics, remoteness, you name it) we need a faster, cheaper and smarter approach than the traditional one which typically goes like: gather the data, send them through networks to the cloud or other environments where they can get processed and leveraged and so forth.

That’s where both edge computing and fog computing really come in. If your data is generated at the edge in IoT, then why not bring all your intelligence and analysis as close to the edge, the source, as possible, with all the obvious benefits. And it’s also where those promised forecasts on edge computing and IoT come in.

So, here are some of those edge computing and IoT predictions:

According to IDC (data announced in its November 1, 2017, worldwide IoT forecasts webcast) by 2020, the IT spend on edge infrastructure will reach up to 18% of the total spend on IoT infrastructure. That spend is driven by the deployment of converged IT and OT systems which reduces the time to value of data collected from their connected devices IDC adds. It’s what we explained and illustrated in a nutshell.

According to a November 1, 2017, announcement regarding research of the edge computing market across hardware, platforms, solutions and applications (smart city, augmented reality, analytics etc.) the global edge computing market is expected to reach USD 6.72 billion by 2022 at a compound annual growth rate of a whopping 35.4 percent.

The major trends responsible for the growth of the market in North America are all too familiar: a growing number of devices and dependency on IoT devices, the need for faster processing, the increase in cloud adoption, and the increase in pressure on networks.

In an October 2018 blog post, Gartner’s Rob van der Meulen said that currently, around 10% of enterprise-generated data is created and processed outside a traditional centralized data center or cloud. By 2022, Gartner predicts this figure will reach 50 percent.

Gartner’s definition of edge computing: “Gartner defines edge computing as solutions that facilitate data processing at or near the source of data generation. For example, in the context of the Internet of Things (IoT), the sources of data generation are usually things with sensors or embedded devices. Edge computing serves as the decentralized extension of the campus networks, cellular networks, data center networks or the cloud.”

What makes edge computing important and different – in a few nutshells

Now a few words on that difference between edge computing and fog computing. We’ll get there.

 
First, fog computing, as term is coined by Cisco as we describe in our article on fog computing. It is sometimes also called fog networking and the word fog refers to the cloud (low-hanging clouds, closer to the edge, right?).

At the same time fog computing is also part of the broader definition and evolution of cloud which IDC calls Cloud 2.0 and encompasses industry clouds and cloud everywhere, including fog.

Edge computing, as a term and an architecture as said exists since longer. However, in the scope of the Industrial IoT edge computing is focused on devices and technologies that are attached to the things in the Internet of Things as this blog post from GE explains. An example of such devices: industrial machines. Fog networks on the other hand focus more on the edge devices that speak to each other, including IoT gateways GE further explains and as you can read below.

As the IoT is all about connecting what was previously unconnected in order to acquire, analyze and leverage data from the assets and devices that contribute to our goals (and still then, a lot of it remains unused) then all the data from connected assets, which could be those industrial machines such as robots, generators, intelligent building components, anything really, we need an architecture to enable this. Both fog computing and edge computing are such architectures with a few essential goals: speed in general and in critical or remote contexts; saving bandwidth, storage, time and costs by limiting the data that needs to be transmitted (as we moved the intelligence to the edge instead and, by definition, decrease network latency).

The difference between the two architectures, fog computing and edge computing, resides in where the intelligence and computing power needed to achieve all those goals sits, as is tackled in this great article where David King of FogHorn Systems and Matt Newton of Opto 22 explain it to author David Greenfield.

In a nutshell, quoting, “edge computing pushes the intelligence, processing power and communication capabilities of an edge gateway or appliance directly into devices like programmable automation controllers” (compare with the GE post). Fog computing, fog networking or fogging on the other hand brings the intelligence to the local area network level and the device or thing, whereby data gets processed in a fog node or in an IoT gateway.

OK, there is more to it than that (both do have their proponents) but hopefully we can now speak about edge computing, knowing that at least we said there is a difference. The main thing to remember? Edge computing is key for IoT and intelligence is shifting to the edge. And data, speed and analytics are important.