One way to track the growth of IT trends is to keep an eye on hiring. When a term becomes a job market, it can have lasting power, and DevOps is a good example.
Type in the term "edge computing" on sites like LinkedIn, Glassdoor or Indeed and you'll see a variety of different IT jobs and titles, most of which don't have the word "edge" in their title.
This makes sense because edge computing is more like a distributed computing architecture than a job title. But as with some important terms before that, such as DevOps and cloud computing. Edge computing seems destined to eventually find its way into more positions, such as "edge architect" or "edge iot developer."
Either way, the edge computing strategy will undoubtedly inform more recruitment plans in the future, meaning new opportunities for IT professionals with the right skills.
"Careers in edge computing are now more common than ever," said Andrew Nelson, chief architect at Insight.
Like cloud computing, edge computing is becoming an industry-transcending IT field -- meaning IT professionals can bring their edge computing skills to a variety of different enterprises. Two major IT trends, the Internet of Things and artificial intelligence/machine learning, are enabling companies in almost any industry to collect, process and analyze data in almost any environment, Nelson noted.
"Any business with multiple locations and data sets can take advantage of these data mining trends and deploy intelligence at the edges for revenue," Nelson said.
Many organizations need IT professionals who are well suited to overcome the specific challenges posed by an edge architecture, such as unique hardware, harsh environments, or intensive security requirements. In fact, edge computing relies on similar skills required in other IT environments and can present additional challenges.
Five skills for edge computing jobs
With that in mind, here are five skill categories that are likely to grow in the edge computing environment of the future. These skill categories can be logically organized, but the specific skills tend to overlap.
(1) Programming, application development, application architecture
While low-code platforms seem to be gaining popularity in fringe environments, this does not eliminate the need for employees with keen software engineering and architecture skills.
"If you want to work on the edge, you have to have good programming skills because you have to optimize your code for specific hardware to run efficiently," said Ryan Ries, head of the data, analytics and machine learning practice at Mission Cloud Services.
This is a fundamental decision on the edge, as it is in the hybrid cloud: Where should the data go? The main appeal of the edge architecture is low latency and the ability to process data as close to where it is collected and used as possible. But that doesn't mean you automatically leave everything on the margins.
As Gordon Haff, a technology evangelist at Red Hat, puts it, "You usually want to centralize if you can, but decentralize when you need to." Businesses with a growing frontier footprint need employees who are good at making these choices.
"Businesses need to think about their data and how they handle it at the edge, whether they try to compress the data and send it all to the cloud, or whether they run the processing at the edge and only send part of the data to the cloud because of possible bandwidth constraints," Ries said.
From a programming perspective, edge environments also require developers to be good at making sure their code works well on unique hardware or under less common conditions.
"Typically, when working on the edge, you might be using dedicated hardware like an FPGA, which requires a lot of code architecture to make everything work," Ries says. This may mean specialized programming skills: people who used to know how to code in Verilog or VHDL have started writing translations in Objective-C to help people who don't know those languages use FPGas."
Building and updating applications for remote sites with resource constraints and specialized infrastructure may require different design or architectural thinking than when developing for a traditional data center or cloud computing environment.
"Edge computing requires a more practical approach to application design, deployment and management, similar to the complexity of infrastructure," says Insight's Nelson. You need efficient data processing and data movement at the edges."
(2) Network and connection
For IT network professionals looking for new challenges, the remote nature of many fringe environments requires networking and connectivity skills to adapt to new requirements.
"Networking is almost always a deployment challenge due to the remote or isolated nature of the edge and the Internet of Things," Nelson said. Solid modern networking skills are required to design and deploy Lans, wireless, and WAN/Internet in a cost-effective and reliable manner."
Wireless networking skills are especially useful, Nelson added, because many edge sites are not radio frequency (RF) friendly. For example, issues such as electronic interference are common in manufacturing and industrial Settings, and there may be some problems with signal coverage in remote environments such as oil fields or agricultural fields.
The specific technologies and protocols that companies will seek will depend on their marginal environment and industry. Proprietary or industry-specific technologies and protocols may be more important in long-established industries, Nelson noted. In any case, emerging networking technologies such as Bluetooth Low Emission, LORA, and ZigBee are also likely to see increasing adoption in marginal areas.
Companies need employees with networking skills because it's important to connect on the edge and ensure that data can be returned to the cloud or data center.
"Without a reliable network connection at a remote site and any central location, data from the edge cannot be profitable," Nelson said.
(3) Infrastructure (computing, storage, data protection)
When the network is good enough, it is good enough -- remote computing and storage is not an urgent need because the data will be returned to a centralized environment. But when low latency is a key requirement, which is one of the fundamental purposes of edge computing, the enterprise will need to be able to deliver the necessary infrastructure resources on site.
For example, AI workloads with large data sets or applications that require near-real-time feedback loops may be better served in the field, meaning they require computing, storage, and other resources to function properly. Nelson points out that edge infrastructure may require capabilities beyond the experience of a data center or cloud computing engineer.
"Managing edge computing at scale is very different from traditional data center management," Nelson said. Running thousands of devices at hundreds of sites with few staff on site can be daunting."
As Red Hat's Haff points out, companies typically can't send out tech support every time these devices need maintenance. Automation and consistency are critical in large edge environments. It's one reason to push standardized technologies and processes from the core or centralized cloud to the edge, says Ishu Verma, technology evangelist at Red Hat.
"This approach allows enterprises to extend best practices from emerging technologies to the margins -- microservices, GitOps, security, and more," Verma said. This enables the management and operation of edge systems using the same processes, tools, and resources as centralized sites or cloud computing."
Low - and zero-touch operations, including the use of digital twining in management, will also become increasingly popular in marginal areas, Nelson said. Storage and data protection at the edge pose similar computing challenges, such as unique power and thermal characteristics, remote management, and specialized hardware.
IT professionals who are willing and able to adapt to the deep data center experience of these and other challenges may find a significant new career opportunity on the edge.
(4) Platform (operating system, virtualization, container)
Some of the same skills (or at least similar skills) that are in high demand in cloud computing environments are also needed in marginal environments, such as containers, choreography, platform engineering, and so on. Standardization and consistency around the edges will again be key.
"When it comes to platform design, the scale of edge deployments demands efficiencies, and it's important to minimize all complexity at the platform level," Nelson said. This includes standardizing on a single operating system, as well as leveraging tools and techniques to achieve consistency and automation across different distributed edge domains.
Containerization, microservices, and infrastructure known as Code (IaC) are all important here. Virtualization is also common at the edges, Nelson said. For consistency and automation, Kubernetes is also a possible choice for edge computing; As Red Hat's Haff has written before, it's not just for server clusters or cloud computing. Moreover, this can be an important link in pushing consistency from the core environment to the edge.
This is expected to be an area where cloud platforms and providers offer more hosting options for running edge workloads.
(5) Edge security
In fact, Nelson and other experts believe that edge security will be Paramount, which means that security professionals will also be in demand.
"The biggest challenge in edge computing is properly securing infrastructure, workloads, and data," Nelson said. "Edge solutions have a much larger attack surface than most other enterprise applications."
Fringe IT professionals need security knowledge and skills even if they are not security practitioners. As with each of the above categories, some adaptation to the details of the marginal environment may be required.
"Security knowledge is a must in the edge computing industry," said Shankar Somasundaram, CEO of Asimily. This can be tricky, and that's because 'edge security skills' require knowledge not only of cybersecurity, but also of embedded operating systems and an understanding of how to optimize limited processing calculations."
Indeed, security at the edge may inevitably require some combination of the above skills, employees who understand threat detection and know how to design and implement remote computing. Somasundaram uses connected cars as a sample use case.
"There is a certain level of safety functionality inside the car, but these processors have very limited processing power," Somasundaram said. Therefore, detecting abnormal behavior inside a car requires a skill set that includes not only an understanding of abnormal detection, but also the ability to narrow it down to a low computational run."
Nelson offered advice on edge security: it is much more effective if built into edge deployments from the start, and much more difficult if retrofitted or added later. This should be familiar to IT leaders who are already running hybrid cloud environments, containerized applications, and so on.
"Security needs to be built into every layer of edge deployment, from the network layer all the way to the application," Nelson said.