07 February 2025

Edge computing offers distinct advantages for enterprises compared to cloud and fog, including enhanced bandwidth efficiency, real-time data processing and analytics, and support for autonomous systems, among others…

“Edge provides much lower latency, often achieving sub-5ms response times, and reduces costs by minimising egress fees associated with transferring data out of the cloud,” shares Anastasia Kundush, Product Director at Gcore. “Security is enhanced by localised data processing, which reduces the risk of breaches and improves compliance with privacy regulations. While many real-world scenarios require some sort of a hybrid approach, pure Edge computing is still preferable for highly latency and security sensitive use-cases.”

According to Eric Simone, ClearBlade founder and CEO, Edge bolsters an enterprise’s security posture by allowing company data to stay local behind company firewalls.

“Edge computing also increases resiliency, as it can run independently if a network becomes unavailable. It can also optimise data usage, as much more data can be streamed locally, enabling more efficient use of real-time data to power AI solutions,” adds Simone.

“The ability to gather data from the field and public areas without compromising the privacy of the individuals is critical,” explains Aaron Allsbrook, ClearBlade co-founder and CTO. “With Edge computing, we can keep all that data gathered in the field, pull out and send back valuable insights, and then delete the original personal data. That’s a huge benefit to Edge computing!”

Who stands to gain?

From industrial automation and smart cities all the way through to transportation and logistics, energy and telecommunications – the industries that stand to gain from Edge are near-endless.

Allsbrook believes that Edge is most beneficial to organisations with field or remote operations and assets: “generally, these businesses have been constrained with how much they can understand, communicate, and optimise what happens in real time. Edge processing allows for bringing IT capabilities like enhanced user guidance, simplified data entry, and AI agents right to the end operators as they do their jobs. It massively shortens the time it takes for business processes to take place.”

“Industries like e-commerce and media that rely on real-time personalisation and instant authentication are prime candidates,” agrees Anastasia Kundush, Product Director at Gcore. “Additionally, organisations in sectors with stringent security and compliance requirements, such as healthcare and finance, gain value from processing sensitive data locally. Edge computing can also greatly extend the functionality of CDNs with features like authentication and personalisation.”
It’s been said that the ‘killer app’ for Edge is AI, meaning that the ability to run AI locally fuelled with massive amounts of private data is extremely beneficial for virtually any business with a physical presence.

“Consider video processing. AI image recognition is powerful, takes a lot of data, is very compute-heavy, and contains sensitive data. Edge computing is an excellent solution here,” says Simone. “Currently, we see Edge implementations in transportation, buildings and campuses, oil and gas, water, and retail. I anticipate significant industry expansion given the flexibility of Edge computing and the power of AI.”

A technical Edge strategy

Picking an effective technical Edge strategy can be challenging, but by considering specific business goals, operational requirements, and technological capabilities, enterprises can make informed decisions.

Organisations should ask themselves: do we need real-time insights from sensors or IoT devices? Is there a need to process large amounts of data on-site to reduce reliance on cloud resources? Are we looking to scale operations without overloading central systems? Is data privacy and compliance a major concern?
“A pilot or proof-of-concept project can test the feasibility of Edge solutions and assess trade-offs before full-scale adoption. Choosing a provider with a robust portfolio of solutions ensures flexibility and alignment with unique needs and is one of the most crucial steps of building an effective technical strategy,” notes Kundush. “Transparent pricing and infrastructure with global coverage are also essential considerations to optimise both cost and performance.”

“An organisation must have a clear and complete understanding of the current technical landscape of its Edge networks,” says Nathan Collins, Regional Vice President EMEA at NetAlly. “This begins with a thorough network assessment that reviews network reliability, bandwidth, and wireless technologies and provides a complete inventory of connected assets and an accurate topology map. This should reveal whether the current Edge infrastructure can handle the expected workloads or identify deficiencies.”

“Businesses should separate Edge hardware from Edge software. First, find software that enables your company to build and configure multiple solutions with the same software,” recommends Simone. “Then for each solution pick the hardware that best fits the job. Finally, you should be able to conduct a pilot that proves it can deliver an ROI in three months or less.”

Edge computing involves moving resources like compute and data out to the network’s Edge to be closer to users, removing latency, and improving application performance. At the same time, those assets must be connected to users locally and back to the central IT network. Looking at who needs access, how they will connect, and how this will be managed over time is an essential practical consideration.

“Additionally, administering, managing, and monitoring these Edge networks for vulnerabilities will require dedicated resources alongside the data and compute resources you have in place. You can’t manage what you can’t see, so strong monitoring and testing resources are essential to make any Edge strategy successful over time,” adds Collins.

Is the network up to the task?

To support Edge computing architecture effectively, the network must be able to handle the unique requirements of decentralised computing, data processing at the Edge, and often real-time communications between devices, sensors, and central systems.

High speeds, bandwidth and low latency are a must-have to enable high throughput for continuous data streams, with high degrees of reliability and availability to meet SLAs. As Edge computing exposes numerous devices, networks, and endpoints to potential security vulnerabilities, data must be protected in transit and at rest, necessitating end-to-end encryption, secure protocols and access controls.

“Network for Edge often comes down to availability and viability and then working backward,” outlines Allsbrook. “This means that your Edge must be smart enough to recognise what network or offline it’s on and then leverage it appropriately. For example, trains and cargo ships regularly go in and out of high- and low-bandwidth areas. The Edge must use compression, priorities, and batching processes to ensure maximum ROI for the overall solution cost.”

“Resiliency is also a consideration; built-in redundancy ensures continuous operation even if parts of the network fail. When the network falters, your Edge solutions should continue to run with the available data and fail ‘gracefully’ (according to your parameters) if the Edge devices run out of storage,” adds Simone.
Meanwhile, Collins believes that implementing and validating proper segmentation and switch provisioning is indispensable to a successful Edge deployment.
“If and when a hacker tries to gain access to the Edge environment, network segmentation restricts their ability to move laterally across the network,” says Collins. “Checking your configuration and segmentation is challenging unless you connect and verify locally, emulating an Edge device or endpoint. Centralised monitoring systems or tools cannot be relied upon to provide complete data about the Edge network. So, you should have both local physical testing as well as overall network level insight available. There’s no substitute for being there to validate that you have the right approach in place and that your critical Edge assets can’t be accessed by those not authorised.”

Getting it right

Before diving into Edge computing, UK businesses should be aware of several potential pitfalls that can affect the successful adoption and implementation of Edge solutions, particularly in the context of the UK’s regulatory environment, technological landscape, and operational needs. Indeed, many businesses underestimate the security challenges posed by the proliferation of Edge devices, instead focusing on securing the core infrastructure.

“The Edge computing model was designed to bring data processing closer to the endpoints that collect it. This delivers performance improvements for applications, but it also expands the potential attack surface area in the Edge,” warns Collins. “You have to consider how you track those assets and ensure that they are secure against vulnerabilities or other security issues so they cannot be used as a jumping-off point to get onto the network as a whole or to attack that particular device for its data.”

Moreover, given that Edge computing often operates in environments where network connectivity is intermittent, if not carefully managed, this can lead to performance issues.

“You must consider the network performance side, particularly when IoT devices are part of your Edge deployment,” says Collins. “Networking issues like Wi-Fi deadspots can affect connectivity and performance, while you will also have to look at your applications’ networking needs and how chatty those applications might be. The proposition for Edge is that you move compute closer to the user, so you don’t want a significant amount of traffic in your application that relies on central processing. That will slow down your Edge application and make it harder for you to achieve your goals.”

Then there’s the matter of scalability: while cloud computing allows businesses to scale quickly and easily by leveraging a central, virtually unlimited resource pool, Edge computing involves physical devices, localised processing, and distributed networks.

“Scaling is very different at the Edge than in the cloud since you are working with a distributed environment,” asserts Simone. “Typically scaling is location-dependent, meaning you add more Edge nodes as you add more locations. For example, if you are monitoring railroad crossings you require an Edge per crossing, or if you are monitoring building energy usage you may need an Edge per building or floor.”

However, Kundush points out that, “for businesses using provider-managed Edge solutions, scalability is typically more seamless than in cloud environments. Edge infrastructure scales automatically without the manual intervention often required in cloud setups. For enterprises managing their own Edge infrastructure, scalability may require additional investment and expertise, but partnering with the right provider simplifies this process significantly.”

UK businesses must carefully consider these pitfalls before diving into Edge computing to avoid costly mistakes and ensure a smooth deployment. It’s crucial to take a strategic, well-informed approach, keeping in mind the unique needs of the business, the complexity of Edge technology, and the long-term goals of the organisation.