Edge computing is a foundational technology for industrial enterprises, offering shorter latencies, robust security, responsive data collection and lower costs.

That's according to recent research from California-based consulting firm Frost & Sullivan, which points out that despite being in its early stage, the multi-access edge computing (MEC) market is estimated to grow 157%, generating revenue of $7.23 billion by 2024 from $64.1 million in 2019. The research also predicts that approximately 90% of industrial enterprises will use edge computing by 2022.

While the effects will be large, the use cases of edge computing are not entirely clear yet. Nor is the definition for many in the industry.

Refining Performance on the Edge

Edge computing, like its name suggests, is the concept of performing computation on devices at the edge of a network, said Raullen Chai, CEO of IoTeX, a Silicon Valley-based technology company that creates an ecosystem for IoT devices. It's like the way that banking transactions now often take place on a smartphone (the edge) rather than at a bank branch (the central server). Data no longer must be transferred from a device to a central server and back to the device, creating low latency that helps artificial intelligence computations work in real time.

“This will be increasingly important as 5G matures, enabling more data to be processed and transmitted than ever before on our smart devices and sensor networks,” Chai said. “With edge computing, the information can be managed at great speed at the drone device itself and operational adjustments can be implemented in real time.”

Edge computing also opens the possibility of providing better security and privacy when combined with other technology. Blockchain stems from the same foundational concept as edge computing: Take the legacy centralized server and distribute its functions directly to many nodes or devices. When combined, edge computing on blockchain-registered devices provides an innovative approach to give users more security, functionality and privacy.

However, it is important to further qualify the definition of edge computing, said Matt Jacobs, chief strategy officer at Fremont, Calif.-based Penguin Computing which builds openLinux-based cloud solutions. Edge computing can be viewed in terms of "far edge," "near edge" and the technology that enables effective interoperation between those environments and larger data center assets. Far edge is the infrastructure deployed in a location farthest from the cloud data centers and closest to the users, whereas near edge is the infrastructure deployed between the far edge and the cloud data centers.

There is demand for the needs of edge computing in targeted markets, such as healthcare and adtech. Edge computing applications can include anything from smart campuses where edge inference for video analysis provides a more secure lifestyle to touchless retail which helps stop the spread of infectious diseases such as COVID-19.

Related Article: The Role of Distributed Cloud Computing in the Enterprise 

Implementation Puts Enterprises on the Cutting Edge

Organizations globally are already starting to move in this direction too and are likely to adopt and apply it to more areas of the digital workplace, according to recent research from Santa Clara, Calif-based Aruba Networks, a wireless networking subsidiary of Hewlett Packard Enterprise.

Learning Opportunities

As networks become increasingly congested with huge volumes of data generated from user and IoT devices, IT leaders are recognizing that analyzing real-time data nearer to the edge yields greater efficiencies and insights, which results in improved business outcomes.

According to the global study of 2,400 IT decision-makers (ITDMs) 72% are already actively using edge technologies to deliver new outcomes, with another 16% planning to do so in the next year. There is also a growing recognition (82%) of the urgency around the need to implement integrated systems to handle data at the edge.

Moreover, the research indicates that the maturity of a company's deployment at the edge is strongly correlated with its ability to derive value from the data collected from devices. Seventy-eight percent of ITDMs in production deployments with edge technologies said they were able to use this data to improve business decisions or processes. That compares with just 42% of ITDMs who are only at the pilot stage and 31% who are planning pilots in the next year. The research also points to three things organization leaders should do when considering moving to edge cloud computing:

  1. Unify: Network operations teams should only consider solutions that can manage all domains and locations from a cloud-native, single pane of glass — a management display console that integrates all parts of an infrastructure — that can centralize and correlate all cross-domain events and operations.
  2. Automate: Network operations teams should only consider solutions that provide reliable, highly accurate and specific AI-powered insights and automation that can resolve issues quickly.
  3. Protect: Network operations teams should consider solutions that use AI to detect, classify and continuously monitor these data-generating devices and work seamlessly with access control.

Related Article: Why Digital Workplace Growth Will Push Cloud Spending in Coming Years 

Use Cases for Edge Computing in the Digital Workplace

Edge computing is the solution to the massive flood of data streaming to and from IoT, mobile and web devices, said Stephen Blum, chief technology officer and co-founder at San Francisco-based PubNub, a real-time communication platform. In executing business logic on the devices as closely to the edge as possible, it reduces traffic sent to external servers and makes it so companies don't have to continuously add data center capacity to deal with growth.

That means better performance (no need to wait for sending and receiving data), lower operational costs and higher security (limiting outward connections). Edge computing takes place as close to the data source as possible, providing a fast and efficient way to act on data. There are five obvious use cases:

  1. Media: Over-the-top media services are increasingly relying on edge computing to enhance their live and on-demand streams. OTT media services need to guarantee a low-latency, high-performance experience not just with their video stream but the additional features as well, from advertising to interactive features complementing the stream, all at massive scale. The bulk of the computing for many of these features can take place on the end-user device so there is no need to send every bit of data back to a central server for processing.
  2. Agriculture: Agriculture relies heavily on edge computing as businesses create more ways to collect data on the performance of farms and factories. Drone flyovers to low-powered sensors to smart machinery create massive amounts of data. They need fast, scalable systems to process and analyze this data as it flows in and new data is created. Farms with tens of thousands of sensors monitoring the performance of the farm (humidity, temperature, water use) do not need to send every new reading back to a central administrator. If the readings are where they need to be, there is no need to stream that data.
  3. Building tiered architecture: Edge computing is a key component of a tiered architecture. It is designed to be near the data source and this allows for direct interaction with applications in a low latency design, said Ryan Mallory, chief operating officer for colocation services at Charlotte, N.C.-based Flexential. The proximity allows for deployment of data directly to end users, driving a better, more connected experience. In today’s always-on, remote culture, real-time nearby data access and application deployment is key to maintaining successful interpersonal connections.
  4. Improving digital experiences: Edge computing is used to drive faster, more secure digital experiences for end-users across industries of all kinds. Data at the edge improves the experience for everything from video downloads, to AI or machine learning, to putting medical data into a doctor’s hands faster.
  5. Infrastructure development: Public/private infrastructure and private 5G manufacturing networks are just starting. Mallory said we will soon see the fruition of the physical infrastructure and the deployment of the virtualized stack, also known as edge compute and application interaction, which is currently in the experimental phase.