Enterprises are pushing cloud resources back to the edge after years of centralization. Even as major incumbents such as Google, Microsoft, and AWS pull more enterprise workloads into massive, centralized hyperscalers, use cases at the edge increasingly require nearby infrastructure—not a long hop to a centralized data center—to take advantage of the torrents of real-time data generated by IoT devices, sensor networks, smart vehicles, and a panoply of newly connected hardware.
Not long ago, the enterprise edge was a physical one. The central data center was typically located in or very near the organization’s headquarters. When organizations sought to expand their reach, they wanted to establish secure, speedy connections to other office locations, such as branches, providing them with fast and reliable access to centralized computing resources. Vendors initially sold MPLS, WAN optimization, and SD-WAN as “branch office solutions,” after all.
Lesson one: Understand your legacy before locking in your future
The networking model that connects centralized cloud resources to the edge via some combination of SD-WAN, MPLS, or 4G reflects a legacy HQ-branch design. However, for use cases such as facial recognition, gaming, or video streaming, old problems are new again. Latency, middle-mile congestion, and the high cost of bandwidth all undermine these real-time edge use cases.
After years of consolidating into clouds, the edge looks different than it did five or 10 years ago. Today, even traditional carpeted offices could be considered the edge, especially if the organization’s mission-critical apps have all migrated to the cloud and its users bring their own devices onto the enterprise network.
This pattern of consolidation followed by decentralization is not a new cycle. The mainframe-to-PCs cycle was followed by the on-premises-to-cloud one, which is now giving way to mobile- and edge-first paradigms.
Current edge technologies are emerging in response to the limitations and expense of cloud computing, according to 451 Research. At the same time, industry-specific use cases, such as retail point-of-sale machines and manufacturing assembly lines, are driving innovation at the edge. Connecting these devices not only streamlines operations, but the data generated from them can deliver insights that help businesses increase their competitive advantages.
Rich Karpinski, principal analyst, IoT for 451 Research, warns that there is no one-size-fits-all approach to edge infrastructure. For both IT (information technology) and OT (operational technology) teams, their first job is to understand their various edge use cases and the expected business outcomes associated with them. Then, they must find the right edge infrastructure and topology to not only get the job done today but to also offer the agility and scalability that will accommodate long-term growth.
“Choose incorrectly, and edge infrastructure becomes a problem in its own right,” Karpinksi warns, “unnecessarily expensive; difficult to deploy, secure, and operate; and limited in its ability to support distributed, modern, AI-driven applications.”
A 451 Research survey of more than 700 enterprises found that:
- 60% of surveyed organizations are reactively investing in their edge, responding to changing IT architectures.
- However, nearly as many (57%) report that their edge investments will also enable them to build new capabilities.
Businesses also expect to achieve an array of outcomes with their edge investments:
- Nearly half (47%) believe edge investments will help them optimize business processes and operations.
- 45% of respondents expect to cut costs via edge infrastructure.
- Meanwhile, 42% believe that edge investments will boost revenues.
- Finally, 39% expect edge investments to reduce organizational risks.
Lesson 2: Pay close attention to what service providers are doing with their edges
Service providers have been investing heavily in edge technologies, both to overcome costly challenges and to pry open new, higher-margin revenue streams. These are still early times for edge clouds, edge data centers, and AI everywhere, but service providers are already placing big bets.
Analysts are also bullish on this market. For instance, research firm IDC predicts that the global market for edge computing will top $260 billion by the end of this year, growing at steady double-digit rate annually through 2028, when the market will reach $380 billion. Research firm Gartner also forecasts fast growth at the edge, expecting the market to hit $511 billion by 2033.
Service provider edge innovation has already arrived in a big way, but its initial benefits start in-house. For instance, as AT&T expands its edge presence deeper into underserved areas, the telecom giant focused on how to drive down expansion costs and eliminate costly, labor-intensive chores, such as truck rolls into remote areas.
According to AT&T CTO Jeremy Legg, AT&T’s global network carried more than 614.6 petabytes of data on an average day in 2022, and the company expects that number to increase by a factor of five by year-end.
For AT&T, solving one problem—the high-cost process of expanding into underserved areas—has unlocked opportunities for new higher-margin services. Case in point: the Cherokee community of Kenwood, Okla. Until recently, tribal members had to drive more than 10 miles just to get mobile cell service. When COVID-19 hit, the lack of connectivity made bad times even worse.
Reliable internet is needed to access tribal services, health care, emergency services, education, and Cherokee language services. With federal funding from the American Rescue Plan, the Cherokee Nation approached AT&T to provide solutions to help close the digital divide and expand connectivity to its citizens.
The Rescue Plan project evolved out of a 2020 collaboration that helped the Cherokee Nation deal with the early stages of the COVID lockdown. The Cherokee Nation, like many other isolated communities, had gaps both in traditional internet and mobile coverage. With the help of AT&T, the Cherokee Nation issued more than 11,000 mobile hotspot devices that were primarily used for virtual learning and telemedicine.
Seeking a more permanent solution, AT&T and the Cherokee Nation sought to bring reliable mobile coverage to Kenwood. In March 2023, they unveiled a 355-foot-tall tower with 5G infrastructure. Having 5G mobile coverage not only provides Kenwood’s 1,000-member population with reliable connectivity, but it also plugs a major public-safety gap by providing Cherokee Nation first responders with dedicated coverage and capacity through the public safety network FirstNet, which supports more than 24,000 public-safety agencies across the country.
“Delivering this dedicated Band 14 spectrum has proven immensely valuable in times when first responders need connectivity the most: wildfires, tornadoes, hurricanes, and countless other natural disasters,” said Jeremy Legg, AT&T’s CTO.
Added Cherokee Nation Principal Chief Chuck Hoskin Jr., “For the first time, citizens here have access to wireless service and more high-speed internet to access the vital resources they need.”
Lesson 3: AI’s impact on networks and computing infrastructure will be broad and unpredictable
Warehouses represent a major bottleneck as logistics teams rely more on digital tools and automation. Manual processes throttle back the promises software vendors make about digital tools. Nokia Bell Labs believed they could overcome this chokepoint by creating digital twins of warehouse facilities using autonomous flying drones.
Nokia’s Autonomous Inventory Monitoring Service (AIMS) uses AI-powered vision processing to navigate the facilities. According to Nokia, the drones scan seven to 10 times faster than humans, and Nokia estimates AIMS could deliver warehouse operators roughly 40% ROI over three years.
The trouble is that an AI-powered, drone-based inventory system requires robust processing, storage, and networking capabilities at the edge.
To accomplish this, Nokia created an edge solution consisting of bare-metal edge servers inside the warehouse that run on Canonical’s Ubuntu Pro and use containerized Kubernetes applications to communicate with drones and update data in the digital twin. In the initial proof-of-concept phase, the build process for each server was manual, as were IT operations. Before they could ship AIMS commercially, Nokia needed to automate key processes and find an over-the-air (OTA) way to scale edge infrastructure to hundreds of warehouse sites without rolling a truck to each one.
Nokia selected Spectro Cloud’s Palette Edge platform to automate its edge deployments and operations. With Palette Edge, Nokia can remotely deploy edge software and centrally manage the lifecycle of distributed AIMS edge Kubernetes clusters.
The first customer for AIMS is Fortune 500 wholesale distribution company Graybar. According to Mark Hirst, Graybar’s vice president of technology, automation is necessary to maintain the health of the industry. Attracting and retaining employees who can deliver accurate inventory counts is becoming increasingly difficult. Graybar believes that not only will AIMS deliver ROI by automating manual, error-prone inventory counts, but it will also help them sidestep a labor crunch.
Lesson 4: Security at the edge is a constant challenge
As the typical organization’s edge attack surface continues to grow, so too does the need for cloud- and edge-native tools that counter new edge threats. London-based McLaren Construction Group had been struggling with several endpoint and network security challenges. The construction firm sought a tool that could automate device authentication and compliance checks, while providing visibility into all networked devices.
“Our IT team was in charge of responding to network access requests, but it took us too much time to handle each request and people struggled with the complexity of integrating with our existing infrastructure,” said Daniel Blackman, McLaren’s head of IT. “We couldn’t control the devices that were connecting to our wireless networks, including IoT devices.”
Lacking control of the edge, McLaren didn’t know what was on its network and it couldn’t be sure that the connected devices were compliant with company policies before they gained access to key resources. Furthermore, McLaren’s IT team struggled to keep up with software updates and security updates. “We were often forced to handle this manually for all devices, which was extremely time-consuming,” Blackman said.
McLaren had an array of mission-critical devices that it lacked visibility into, including CCTV’s, audio conferencing devices, security biometric readers, and video conferencing equipment. Blackman and his team narrowed their search to cloud-based network access control (NAC) providers that could deliver edge security with no resource overhead and no added complexity, particularly in terms of introducing new hardware or infrastructure changes. Eventually, McLaren selected the SaaS NAC service from startup Portnox.
Now, devices must connect through Portnox CLEAR, a cloud-native NAC service. CLEAR controls access to the network based on the 802.1X protocol. It can block rogue devices, quarantine non-compliant endpoints or limit access to specified resources using access control lists (ACLs) or VLAN changes. Users must connect through a CLEAR portal, which assists with connecting to the network and manages the remediation of devices, showing users the steps to return a device to a healthy security state.
In addition to device authentication and control, CLEAR further secures McLaren’s edge through rogue device detection, automated policy enforcement, compliance checks, automated patching, and ongoing endpoint risk monitoring. According to Blackman, his team had CLEAR set up within 30 minutes, and it’s now easy for them to scale security to other edge locations.
Lesson 5: The enterprise edge will continue to sprawl indefinitely
The enterprise edge will continue to sprawl, even for organizations that wish that were not the case. With every device more complicated than a butter knife, connectivity increasingly comes standard, be it your office coffee maker or an expensive piece of machinery on your factory floor.
A good example of this is electrical infrastructure. Back in April 2024, GE spun off its energy manufacturing business line, GE Vernova (NYSE: GEV). “GE Vernova is purpose-built to electrify and decarbonize the world,” said CEO Scott Strazik on the day the spinoff was completed.
Since the spin off, the company has positioned itself to develop sustainable power sources for emerging technologies, such as AI and IoT, and it has committed to investing nearly $600 million in its U.S. factories and facilities over the next two years to help meet the surging electricity demands around the world.
GE Vernova makes all sorts of electrification equipment, from natural gas turbines for power plants to smart-grid infrastructure to hydro-power equipment. As GEV modernizes power generation, it is also finding multiple ways to bring connectivity to its equipment at the edge.
GEV is partnering with EnergyHub to bring smart-grid capabilities to a range of devices, including smart thermostats, electric vehicles, solar panels, and batteries. GEV is also working with Anterix, Ericsson, and Nokia to provide bundled services to utilities over 900 MHz private wireless networks. The collaboration focuses on enhancing grid communications infrastructure and driving digital transformation within utilities.
What GEV, McLaren, AT&T, Nokia, and many, many others have learned from their early forays into the expanding, hyper-connected, AI-influenced edge is that the edge will sprawl across the economy in a variety of ways, whether the growth is planned or not. Finding ways to remotely maintain, update, and manage edge infrastructure in a cost-effective, scalable way is just the beginning of the shift of resources back to the edge, a shift that enterprises must prepare for one way or another because service providers, manufacturers, and even your own employees will continue to extend the boundaries of the enterprise edge.
Source:: Network World