IoT

Digital Transformation at the Edge

Virtually every major company is racing to transform digitally, with a real sense of urgency. In their rush, they risk overlooking the need to modernize and innovate at the edges, areas expected to generate growth for many industries.

Microsoft CEO Satya Nadella predicts the world will have 50 billion connected devices by 2030, changing business thinking and how data is used. Forrester projected that the edge computing market will grow by 50% in 2020.

These developments require businesses to look beyond their traditional understanding of the edge. A growing number of companies see that the flood of new data — gathered and managed at the edge — will be crucial. This information will drive efficiency, and it will provide real-time customer analysis that allows companies to make smarter decisions and potentially create new business models.

A Gartner survey showed that two-thirds of organizations intend to use 5G networks to bolster their edge computing capabilities. The research and advisory firm predicts that edge computing infrastructure will be completely deployed between 2025 and 2030 to support this digital transformation.

The results will help companies create data-driven organizations that operate like a live enterprise, one that adapts to new environments. But without modernizing that critical slice of their network, their digital transformation will be incomplete.

Edge challenges

Organizations cannot modernize the edge without taking a close look at its network infrastructure. A robust, flexible, and agile backbone is necessary to roll out new applications worldwide to a diverse group. A failure here will lead to poor user experiences and an eventual loss of business.

The current edge uses traditional wide area networking (WAN) solutions. This hub-and-spoke model transports traffic from the edges to centralized data centers. This approach worked in the past, but it fails now when it comes to cost, speed, and performance. Our clients have found that WANs are likely to suffer from:

• Low agility — These systems are less flexible and are unable to automate a new application rollout with a single click. WAN systems also can’t quickly reprioritize application traffic as business needs change.
• Slow time to market — Traditional networks are unable to quickly provision and configure edge devices and applications and then connect them to central data centers. Doing so can take weeks or even months with legacy technology.
• Poor control — WAN is unable to control the bandwidth utilization based on consumption demands. Recurring bandwidth cost claims a big chunk of IT budgets.

Traditionally, edge infrastructure has acted as either a transaction point or a data collection point. For most processing, the edge communicated to the central data center relatively efficiently. Not much was expected or needed from the processing power in these distributed devices.

However, as data collection grows exponentially, the previous generation of technology is showing its age. We have devices that monitor manufacturing equipment on factory floors and internet-connected video cameras sending live footage from retail stores. There is a risk that bandwidth won’t keep up with the volume of data transmitted. This will increase latency and, as a result, provide customers a worse experience. Also, this affects the bottom line directly through higher bandwidth costs.

The new edge

While the old approach to edge computing is no longer feasible, fresh thinking exists and viable solutions are now within reach. Technology is readily available to enable the edge to:

• Centrally manage and automate functions in order to deliver the needed speed, agility, and intelligence.
• Dynamically determine the priority of network resources, path selection, and policies in a highly adaptive way based on the importance of business applications.
• Allow the handling of traffic routing across multiple transport options to multiple destinations, such as a software as a service provider, public cloud, or private cloud. But that can’t come at the expense of robust network security at every ingress and egress point in an automated, transparent way.
• Perform artificial intelligence and analytics tasks and quickly engage users in a smart way to deliver a superior experience. That approach can include interactive bots, self-healing, and predictive intelligence.

Technology transformation at the edge

Over the years, both networking and computing technologies have evolved in ways that support the needed edge transformation. Here are some of the technologies our clients are using to make this new approach possible:

• Software-defined WAN — This has emerged as an important solution that can eliminate the flaws of traditional WANs. These are application-aware, agile and secure networks that can be provisioned, orchestrated, and managed from a central console. The technology uses a software-defined overlay to create transport-independent and policy-based routing with faster implementation times. The multiple transport options reduce cost while enhancing reliability, performance, and experience.
• Computing power — A use of on-premises software or cloud-based edge computing technology brings greater computational power and data storage closer to the devices. This allows them to perform analytics at the edges while containers and other technology add orchestration and automation. These advances lower latency and inevitably improve performance, especially when it comes to demanding data transfers. They also reduce bandwidth expenses and eliminate the need for some high-end hardware platforms.

In tandem, these new approaches to edge computing are not just beneficial but inevitable. Users demand a better experience and more engagement, while businesses need real-time analytics. Companies focused on advancing their business models will look to the edge to see what is possible.