Edge computing is forcing businesses to rethink their approaches to IT infrastructure with its new and exciting approach to network architecture that’s breaking beyond the limitations of cloud computing. But is edge computing little more than a buzzword? Or does it represent a complete sea change for every industry currently relying on cloud computing?

Edge Computing

Edge computing is an industry buzz phrase (much like “The Internet of Things”) that essentially means that the computing of data is happening that much closer to where the data is actually being created – the source.

Cloud computing is constantly evolving, with both businesses and personal users now accessing centralised services on a daily basis – whether it’s platforms like Dropbox to share files or even email services like Gmail. Even our entertainment devices and platforms (Netflix, for example) are powered by cloud computing and the vast majority of that computing power relies on only a few major public cloud providers.

The big four are Amazon (Amazon Web Services, or AWS), Microsoft (Azure), IBM and Google, with AWS boasting a market share just north of 34%. The advent of more readily available 5G will also allow consumers and businesses faster, more stable access to their cloud-based data and applications without having to use Wi-Fi. The problem with the current cloud situation, however, is that there isn’t much room left to grow and all the opportunities lie at the edge – closer to the source of the data.

According to Marco Argenti, vice president of IoT at Amazon Web Services, there are three “laws” which call for edge computing. While these are not concrete laws, they do work as a rough guide for figuring out just what an edge computing solution could bring to your enterprise – let’s take a closer look at them below.

Physics – When you are operating in sectors that require split-second transmission, but are physically far away from a hub, you need to be closer to the action with short latency. For example, service providers in industries such as on-site construction management, emergency support services (paramedics) and remote logistics management.

Economics – Currently, it might not be economical to transfer large quantities of data from the edge to the cloud; it might make more sense to pre-process it locally. Soon, with the advent of widespread 5G connectivity, this might no longer be the case.

Land – There might be reasons you want to keep your data local and filter it before it regresses to the cloud. A business might have compliancy requirements such as regulations within the financial industry, whereby certain data needs to stay local which is a term that in itself is prone to subjectivity.

The Benefits

The primary benefit of being so close to the data is the extra speed since edge computing processes data in data centres closer to the source. This means significantly less latency as the data doesn’t have to travel as far. Data is, after all, bound by the speed of light and with the world expected to be generating 44 zettabytes of data by 2020, the less mileage that data has to travel, the better the results for the end user. This is particularly beneficial for large organisations that are processing large quantities of complicated data daily.

Besides the more obvious speed benefits, edge computing also provides security advantages. Because edge computing distributes its power across a range of devices and data centres, it makes it that much harder to bring down the network. The distributed nature of edge computing architecture makes it easier to implement security protocols that can seal off compromised portions without shutting everything down.

As data is first transmitted from many sources and then recombined before executing, you might assume there is a time-lag penalty, but edge computing makes sure there is no disconnect in real-time information processing.

Alongside the security advantages, it also offers greater reliability, with less chance of network problems in far-flung locations affecting the data closer to home. Even in the event of a local data centre outage, most local IoT devices will be able to operate independently since data can be rerouted to ensure end users retain access to their information.

Edge computing also shares the scalability of traditional cloud computing without many of the associated pitfalls, and it’s significantly more versatile – with the capability to give content providers access to reliable uninterrupted streaming.

Edge computing is always on, always connected and always gathering data too. The unstructured information gathered by edge networks can be processed locally or delivered back to the network core, where it can be dissected by machine learning to identify trends data points.

Who Should Be Using It?

There are applications, particularly in situations where complicated IoT devices are being used more commonly in remote locations – where there simply isn’t enough bandwidth to send the vast amounts of data required in a reasonable timeframe using conventional public cloud services.

It could also substantially benefit the manufacturing sector by improving production line productivity and lowering costs to manage ‘m2m’ devices. Where edge computing could make a really significant impact, however, is the healthcare sector, where vast quantities of time-sensitive data are being processed on a daily basis. According to Richard Corbridge, the chief information officer at Leeds Teaching Hospitals NHS Trust, it might be the answer to a lot of digital healthcare problems. He explains:

“We have a basic pilot of edge computing underway; in the renal ward of our hospital we collect data from IoT devices that are monitoring patients, we transfer this information directly into the trust’s electronic health record from bedside. The edge element allows national early-warning scores for patient care to be analysed, compared to the information during the length of stay of the patient and, where necessary, alerts or tasks are created.”

Who’s Using it Now?

In the US, 21% of companies are already using the technology and while the UK appears to be lagging behind its transatlantic cousin with only 9% of companies currently using it, 32% of companies are estimated to be using it within the next five years.

As leaders of the cutting-edge, Google was one of the first companies to experiment with ‘edgy’ applications and practices. Their Progressive Web Apps, for example, have ‘offline first’ functionality, allowing you to do work locally before syncing with the cloud.

Google Clips, meanwhile, keeps data local by default. Google is also one of the many major tech companies investing heavily in the future of self-driving vehicles, which arguably take edge computing to its logical conclusion, as autonomous cars can use up to 4,000GB of data in just one hour of driving, requiring all the bandwidth it can get!

In the US, the Sprint telecommunications company’s new IoT service “Curiosity” is using edge computing services supplied by edge computing provider Packet, whose CTO, Ihab Tarazi, feels that wireless companies taking advantage of the technology are just the beginning. He says:

“There are big enterprises that have a lot of stuff in the cloud, but because they want to improve the performance of the apps, or mobile payments or something, they want to deploy something at the edge. It’s already happening.”

Sprint also used CES 2019 to announce the development of its own ‘smart-city’, which simply wouldn’t be possible without the speed and versatility of edge computing.

The Risks

With any new major technology, it’s always natural to ask what we’re giving up in order to make these leaps. For the consumer, the major drawback is that the companies using edge computing for IoT purposes will, by default, have more control over more of their life experiences and it will take some of the agency out of the consumer life. But for many users, it might be something of a relief to take their hands off the wheel somewhat.

For businesses, the drawbacks and the risks are few and far between. It does require more local hardware if you’re going to take advantage of the full IoT capabilities and more connected devices mean more opportunities for cyberattacks. Finally, as edge computing only processes and analyses a subset of data (discarding raw information and incomplete insights), it will be up to businesses to consider what level of information loss is acceptable.

Getting Started

Today’s data centre colocation providers can offer the best means for filling the gap in IoT’s edge computing landscape, whilst also offering a cost-effective means for managing, storing, and organising big data. Colocation is the most efficient and flexible means of managing and analysing the enormous amounts of IoT sensor data for factories, supply chains, power grids, distributed products and even ‘smart cities.’

This coupled with key partnerships positioned at the data centres, such as ISVs (Impartial Software Vendors), Market/Data Analysis Service providers, and a vast list of NSPs and MSP, configuring an edge network at a leading data centre would be considered a ‘best execution venue.’

Telehouse’s UK data centres boast connectivity to over 530 mobile and network connectivity partners and as the UK data centres are located in the centre of London (at the famous Docklands Campus), they are perfectly placed for London-based businesses wishing to set up edge-based ICT infrastructure.

The Future

With more connected devices hitting the market every day, fuelled by IoT and 5G development, enterprises that have thus far only scratched the surface of what edge computing has to offer will no doubt be clamouring to invest more into edge computing and blaze a path for the future of the cloud.

Edge computing offers many significant advantages over more traditional network architecture and it doesn’t have to cost much more either. Gartner predicts that by 2022, over 50% of industrial IoT analytics will be performed at the edge. So why not get ahead of the pack today?

In the next few years, there will be a significant need for enterprises of all sizes to empower edge computing. If your enterprise collects data from the periphery of the cloud and you find yourself drowning in too much data, or if you’re unable to provide the fast and reliable results that your users expect, it might be time to review the benefits and consider how to configure your infrastructure towards this bold new decentralised technology.