Edge Computing Deployment: 4 Key Considerations

By 2028, organizations will have spent $700 billion on edge equipment and facilities. The need to reduce latency in real-time applications in IoT and mobile computing is the primary driver of investments. Over time, however, edge computing will support a wide range of applications in business, industrial, and consumer environments. Any use case that requires real-time speed for data collection and analysis is a candidate for edge computing.

Like any emerging technology, edge computing is a work in progress. Edge relies on a different networking model with data collection and processing done at edge nodes instead of at central data centers or cloud applications. Deploying and managing distributed edge servers and smart devices presents challenges to edge computing providers and customers.

Four primary edge computing deployment challenges

The four primary challenges involve connectivity, staffing, security, and maintenance.

Connectivity

Today, just 10% of enterprise data is created and processed outside a traditional centralized data center or cloud, according to Gartner. By 2025, however, Gartner expects that to rise to 75%, due to the proliferation of IoT devices in manufacturing, oil and gas, healthcare, security, transportation, and other industries. This will vastly increase demand for bandwidth, especially at the network edges.

Data center providers need to adapt their infrastructure and services to support the needs of different edge computing applications. For example, an energy company that collects and analyzes near-real-time data from customer devices needs a network latency of no more than 10 milliseconds. To achieve that, edge servers or edge data centers must be in each customer market. Alternatively, a real-time application such as robot-assisted remote surgery or self-driving cars require even lower latency. Depending on the distribution of the robots/cars/devices, that might mean putting micro edge data centers in dozens of office buildings or bus shelters around the city, or hundreds embedded in traffic lights.

Also read: Edge Data Centers Connect You with End-Users Faster

Wireless network providers will also have to adapt to this flood of smart devices. They’ll need to expand the number of cellular sites to avoid congestion as well as deploy 5G wireless networks to speed response times for mobile users and IoT devices.

Skilled staff

A major challenge for many companies is finding staff with sufficient understanding of edge computing and the skillsets needed to implement and manage it. In a survey of process, plant, and production engineers, for example, 46% said that a lack of knowledge about edge technology and applications was the biggest barrier to edge deployment, and 30% reported a lack of in-house expertise needed for edge computing.

Edge planning and deployment requires a mix of skills that may be difficult to find in some markets. For example, the surveyed engineers listed required skills such as process or automation engineering, computer networking, cloud computing, cybersecurity, data engineering, data science, and application development. For the majority of organizations, the best path forward is to get help from IT consultants, edge data center providers, and managed services providers.

Security

Edge computing is often perceived as a security threat by IT teams. That belief stems, in part, because of the lack of security controls built into most IoT devices. That vulnerability is multiplied by the number of devices connected to the network. One compromised device opens the door to other edge devices and to the network as a whole. That can expose customer data and other confidential information. Edge devices need better security, starting with password management. Many customers don’t bother to change the factory default passwords on their devices. Edge network equipment should be physically secured as well as monitored with network management software.

Gartner recommends security including data-at-rest encryption, boot-time integrity checks on devices, security controls on software and software updates, and zero trust network access.

Management

Managing remote edge servers and hundreds or thousands of connected devices is a major challenge, especially for IT departments accustomed to central data centers with on-site IT staff.

Two solutions are remote management and automation. Rules-based automation can keep devices operating without the need for constant human oversight. IoT management applications, such as the ones used for manufacturing devices or transportation fleet management, usually provide remote management and automation capabilities. Cloud and edge platform providers also offer remote device management. For example, Microsoft Azure’s IoT automatic device management provides tools for deploying and updating groups of devices.

Even automation and remote management may not be enough in some cases. Equipment eventually breaks, and it rarely does so on a regular schedule. Data center service providers are increasingly offering managed services for edge deployments. Few organizations can afford to station an IT employee at every remote edge server. Organizations should look for managed edge services and “remote hands” support services to help handle maintenance and management.

Data center providers, colocation facilities, and hyperscale cloud providers all provide different types of edge services, so it pays to study your options. For instance, an organization that wants to own and control its edge server, but not an entire edge facility might be well served by a colocation facility with shared rack space, power, cooling, and security. Another customer may not want to own or manage any edge equipment. In that case, a data center services or hyperscale cloud provider may be the best option.

A cloud provider can help you deploy edge computing

Different providers can offer different types of connectivity, bandwidth, minimum latencies, data storage, and managed services. TierPoint provides edge services through its geographically diverse mix of 40 world-class data centers in 20 U.S. markets, all connected by a coast-to-coast network. TierPoint’s data center fabric provides any-to-any high-speed connectivity with advanced networking technologies and a menu of managed services. We can help you with your edge deployment. Contact us today to learn how.