Internet of Business says

For over a decade, consumer and enterprise technology users have had their heads in the cloud. But with the ever-growing workloads that will be tied to 5G, autonomy, the Internet of Things (IoT), and real-time analytics, businesses are starting to look elsewhere for their computing needs.

We are connecting more devices, and crunching more data more quickly than ever before. The latency of hosted solutions demands a new approach, especially for time-critical processing – in some smart city or connected/driverless transport applications, for example, where split-second reactions may be essential.

As a result, users and providers alike are moving towards the so-called edge environment, with vendors such as Dell and Microsoft investing billions of dollars in IoT portfolios and edge computing services, which move processing and intelligence closer to the point of need.

Speaking in New York at the launch of Dell’s IoT division last October, founder and CEO Michael Dell said, “The edge will be everywhere and everything: that is the Internet of Things and, ultimately, it will be the Internet of Everything.

“With the cost of a connected node approaching zero dollars, the number of them is exploding. We’ll soon have 100 billion connected devices, and then a trillion, and we will be awash in rich data.”

Dell’s strategy now is to create the distributed infrastructure to move customers’ innovations into a continuous cycle of renewal, he said, with the edge enabling real-time analytics on real events.

The enterprise shift

The current wave of enterprise cloud computing has long been shifting processing off premise and compelling us to put ‘a lot of the same things in a few places’, namely huge third-party data centres – many of which may be shared and/or offshore. (It’s wise to remember that ‘the cloud’ has nothing to do with a fog of borderless code floating in the sky, and everything to do with server farms built on land under national data hosting, privacy, and transfer laws.)

With the rise of the IoT and embedded intelligence, the trend is moving the other way for many applications, putting ‘a few things in a lot of places’, which is a totally different strategic and operational technology challenge.

Put another way, the pendulum of the IT industry is swinging away from centralised services – such as those in the cloud – and back towards distributed systems, especially for real-time data processing in IoT, sensor array, AI, data triage, and machine-to-machine instant messaging applications.

Zac Smith, CEO of bare-metal server specialist Packet, believes this shift requires a different approach to infrastructure – a more open and agnostic one that can embrace variety and rapid changes in hardware.

“History shows us that as workloads grow in size or importance, you start to see the hardware become much more specialised,” he says. “Add the impact of latency and regulation to these kinds of experiences, and you start to see a very different infrastructure requirement than what exists today.”

The advantages of edge computing

These increasing data and real-time analysis requirements have given rise to edge computing, and other distributed approaches. By placing data storage and analysis capabilities at the edge – as close as possible to the sensors, pumps, generators, or whatever hardware is crucial to your operations – users are given several advantages over cloud-based solutions.

Reducing the amount of data sent to the cloud cuts out unnecessary data transfers, simplifies cybersecurity, and decreases network and system response times, meaning data and analysis on critical processes are kept as up to date as possible.

In short, cutting out the latency of a cloud-based network could mean the difference between addressing mission-critical issues just in time, or getting information that is too little, too late.

Cloud vs edge

Of course, rarely in the IoT space does one size fit all. Cloud solutions are often cheaper, more powerful, easier to implement, integrate, and scale, and shift the cost burden from capital to operational expenditure. Plus, there are many advantages in storing data centrally and off-premise when it comes to mobility, remote working, collaboration, and flexibility.

Ultimately, it boils down to workload urgency and costs. How much are users willing to process their workloads on scarce resources at the edge? And which elements can be handled in off hours, or shipped off via low-orbit satellites to a location with cheaper power?

Workloads that aren’t latency specific will continue to be served from the cheapest reasonable locations. Most organisations would struggle to match the economies of scale achievable by the leading cloud providers, such as AWS, Google Cloud, Microsoft Azure and IBM, when hosting data processes locally.

Meanwhile, the recent generation of enterprise collaboration platform providers, such as Box,, and others, provide integrated cloud solutions that users find attractive and easy to use.

A combination of both cloud and edge computing suits many organisations. And, as these and other hybrid approaches proliferate, supporting software will become ever smarter, helping decide which workload gets handled where, according to business strategy, regulations, economics, and more.

Some businesses are also using the networking architecture known as fog computing, whereby instead of collecting and analysing their data in ‘the’ cloud, or at the edge, it is hosted on a local area network – a halfway house between the responsiveness of edge computing and the centralisation of cloud computing. The manageability advantages of this alternative approach are certainly attractive to some IT leaders.

But as 5G becomes commonplace over the next few years, and early use cases start to accelerate, edge computing is likely to gain significant traction.

Operating at the edge

If there was any doubt about the increasing popularity of edge computing, you only need to look at the influx of major announcements and partnerships in the space for confirmation.

Two early adopter of edge services for the IoT, Dell Technologies and Microsoft, recently joined forces to create an integrated Internet of Things (IoT) offering to help vertical customers simplify IoT management, and enhance security between the edge and the cloud.

The new hardware and software platform unites Microsoft’s Azure IoT Edge offering with Dell edge gateways and VMWare’s Pulse IoT Center. This enables centralised monitoring and helps lower the cost of running IoT networks at the edge.

Microsoft CEO Satya Nadella acknowledged the company’s burgeoning edge business in its latest financial results:

“Our early investments in the intelligent cloud and intelligent edge are paying off, and we will continue to expand our reach in large and growing markets with differentiated innovation,” he told Wall Street analysts.

“We reorganised our engineering teams to break free of the categories of the past and better align with the emerging tech stack from silicon to AI to experiences, to better serve the needs of our customers today and long into the future. And, most importantly, we drove innovation to deliver differentiated value across the cloud and the edge.”

After a limited launch last year, Microsoft announced the global enterprise availability of its Azure IoT Edge technology in June. The platform is open source (and available on the recently-acquired GitHub).

Microsoft is also creating an ecosystem of certified hardware and software for the edge, called the Azure Certified for IoT programme, which will certify core edge functionalities, such as device management and security, and serve as a place where developers can find pre-built edge modules (now available through Azure Marketplace) to accelerate edge solution development.

Early customers of Azure IoT Edge have praised the technology’s ability to enable increased uptime and offer real-time insights on the performance of operating equipment, thanks to intelligent applications that run right on the device, and in locations with unreliable connectivity.

Vulcan Steel, for example, is using AI and Azure IoT Edge to review thousands of pieces of video footage a day and highlight any risky behaviour that could cause an accident in the loading and unloading of the company’s trucks.

The industrial angle

The Industrial Internet of Things (IIoT) will be a hotbed of edge computing applications.

Certainly, chip giant Qualcomm sees such a future, and recently announced its new Wireless Edge Services offering. The suite of software services is aimed at helping both enterprise and IIoT customers to provision, connect, and manage intelligent wireless devices securely through their cloud platforms.

With the proliferation of huge numbers of connected devices, many of which need to carry out analytics at the edge, a centralised means of managing them is essential.

Google finds its edge

Google was late to the party but has also staked its claim on the edge landscape. After unveiling its enterprise Cloud Services Platform, it recently launched two new products aimed at helping customers develop and deploy intelligent connected devices at scale.

Edge TPU is a new IoT-focused chip for the edge environment, and Cloud IoT Edge is a software stack that extends Google Cloud’s AI capability to gateways and connected devices. Together, the two products allow users to build and train machine learning models in the cloud, then run them on Cloud IoT Edge devices via the Edge TPU hardware accelerator.

Cloud IoT Edge extends Google Cloud’s data processing and machine learning capabilities to gateways, cameras, and end devices, making IoT applications smarter, more secure, and more reliable, according to Google.

Google also announced an on-premise version of its new enterprise offering: as clear an indication as any that the cloud is no longer the only option.

Is edge computing for everyone?

However, not every sector needs to rush to the edge, as Laz Vekiarides, CTO and co-founder of enterprise hybrid cloud storage provider ClearSky Data, explained in our May interview:

Edge computing is based on location of data, location of users, and performance demands, so companies need to decide how much data they need locally – at the edge – to process and make decisions in real time.

“Anything that doesn’t require real-time interactions can potentially run smoothly without edge computing. It really depends on how responsive your device needs to be, and whether or not it can live with the delays of going to and from the cloud.

“For example, I don’t need my smart thermostat to respond to me instantaneously. I can wait an extra 30 milliseconds for it to complete my request.”

The edge of the city

However, one area of the IoT that really stands to benefit from edge computing is the smart city. In May, Edge intelligence startup SWIM.AI announced new smart city and Internet of Things (IoT) offerings, powered by its EDX AI software, as well as edge products for smart grids in the utilities, oil, and gas sectors.

Earlier this year, the 2015-founded company came out of stealth mode to launch its EDX platform. The product is designed to supply business insights in real time from IoT devices, by autonomously building digital twins from streaming data at the edge, via real-time analytics and predictive machine learning.

SWIM EDX for Smart Cities integrates with existing urban infrastructures to gather and interpret local data. The product can combine city, traffic, vehicle, pedestrian, parking, and sensor data to power real-time smart city applications.

Driving on the edge

Within connected cities, autonomous and connected vehicles will become ubiquitous, and in many ways they represent a microcosm of the benefits of edge computing.

When an autonomous vehicle needs to avoid a collision, the edge environment and the distributed core are where the real number-crunching has to take place. A self-driving car needs to be able to respond near-instantly to the reams of data being reported by its LiDAR, sensors, camera equipment, and other systems, and then processed by its onboard computer.

An autonomous vehicle is essentially a network on wheels, including edge devices, networking equipment and a central computer.

But such vehicles will also require cloud connectivity, to update navigation and traffic data, enable automatic payments at service stations, and other smart onboard services, such as entertainment.

So, as we start to see more smart devices, drones, AI systems, and machine intelligence taking hold, we’ll see a future in which the edge takes on more and more computing workloads. And as more endpoints become significant generators of data, and as 5G networks expand, edge computing will come to the forefront.

Yet cloud computing will always play a significant role in the enterprise, particularly as hybrid computing networks become better able to leverage edge technology where suitable and, likewise, utilise the benefits of the cloud for less urgent workloads.

Ultimately, it’s a case of recognising that cloud computing is no longer the answer to all of our data processing and storage needs. The edge is leaving its mark.

Additional reporting and analysis: Chris Middleton

To learn more about implementing cloud and edge technology visit IoTBuild, London, on 13-14 November 2018

IoTBuild network to edge