17 Aug The Hitchhiker’s Guide To Edge Computing
The topic of edge computing is hot. Every analyst, journalist, and pundit is offering an opinion. Every day, my email “inbox” is filled with enough of those opinions to thoroughly confuse anyone trying to get their head around the basics of edge computing and what all the hype means. So, in a thousand words or so, I’m going to try and distill some concepts around edge computing and explain why all of this matters to you—whether you’re an IT practitioner, an industrial automation engineer, or just a reader trying to get better educated on the topic.
What is edge computing?
Edge computing is a model where IT infrastructure and applications are moved beyond the physical datacenter. Think of a manufacturing company with factories across the world. By deploying servers, storage, and analytics software to the factory floor, the organization can collect, scrub, and analyze factory data locally, allowing a greater level of control over each assembly line.
Take this model and apply it to a wastewater treatment plant, or a large hospital. Then, apply machine learning and artificial intelligence to the equation. Get the picture? This is why edge computing is such a hot topic.
While the answer is seemingly yes, it’s actually no. Automated assembly lines on a factory floor are nothing new, it’s true. But now we’re talking about the collection and real-time analysis of raw data, leading to actionable intelligence that can optimize the efficiency of a factory’s assembly lines for real cost savings. Alternately, consider a smart traffic grid that can automatically reroute traffic based on traffic or an accident. This isn’t science fiction—it’s edge computing.
Like the early days of the internet or cloud computing, there is a lot of hype around edge computing. However, it’s not just hype—the adoption of edge is accelerating at a fast pace. The reason for this has to do with the fact that sensors, machines, and devices (the Internet of Things) are already deployed in many organizations. Data is generated, collected, and used to manage and control industrial systems. The path to business and operational efficiency is within reach, and the savings associated with these efficiencies makes deploying an edge strategy worth the investment.
OT versus IT – it’s a bigger issue than you think
Operational Technology (OT) is the hardware and software designed to manage physical devices. Think of a wastewater treatment plant that requires valves to be opened and closed in specific sequence. OT refers to the control systems that would be used to create and manage this environment (see SCADA, PLC, PXI).
There are two challenges that organizations face when planning edge deployments:
- Technical. The systems, software, and networks used to manage industrial environments tend to be proprietary and closed, and each industry is different, leading to some level of fragmentation. Integrating with traditional IT systems that can apply analysis and intelligence can be difficult.
- Cultural. OT deployments are usually managed by engineers, plant operators, or other professionals who, though very technical, are not well versed in Information Technology. Conversely, IT professionals typically have little to no experience in OT. This dynamic can lead to a situation where plant operators are unable to support expensive IT gear sitting on their plant floors. Alternately, IT organizations could be collecting mounds of data without the ability to make any sense of it.
So which challenge is bigger? It’s difficult to say, since every situation is different. Organizations planning an edge deployment would be well served to reserve a seat at the planning table for both IT and OT organizations. Equally important is the development of a comprehensive plan of attack that builds cross-organization training and awareness. Edge deployments are often installed in facilities that lack proper IT support. In these situations, IT organizations need to ensure that plant operators and engineers are fully trained and capable of supporting the servers, storage, and networking gear on location.
The technical challenges associated with merging IT and OT functions can be equally daunting. The good news is that server vendors (such as Dell Technologies and Hewlett Packard Enterprise ) and vertically aligned ecosystem partners have built practices in this area (more on this in a bit).
The edge ecosystem
The edge ecosystem varies depending on the vertical. For example, the companies, solutions, and standards that support the healthcare industry are different than the ecosystem that supports manufacturing (which is different from industrial automation). Companies like Schneider Electrichave built practices across multiple verticals, and this breadth is often achieved through strategic acquisition (market consolidation).
One of the more interesting players in the edge ecosystem is Stratus Technologies. Those familiar with Stratus understand the value Stratus brings to the enterprise with its highly fault-tolerant server solutions. Stratus built a strong practice in the financial services industry and recently introduced its ztC Edge (you can read my coverage here).
What makes Stratus interesting is the depth of its ecosystem. It tightly partners up and down the solution stack to deliver into companies across many verticals. A good example of this is AutomaTech, an integrator based in the U.S. that focuses on delivering SCADA & automation into industries such as manufacturing, wastewater, and power & utilities.
The breadth of specialized partners and the depth of those partnerships are serving Stratus well.
What about the server vendors?
In a market like the edge that can get so specialized, it is easy to get really deep, really quickly and forget about the likes of Hewlett Packard Enterprise and Dell Technologies. Both HPE and Dell have built entire practices around edge computing, and each offers unique solutions with value propositions that position them well in this market. I’ve been impressed by both companies’ understanding of how specialized the edge market can be, and how they’re developing accordingly the relationships and strategic partnerships necessary for success.
Maybe a better question to ask is, why the server vendors would play in a market that can be so specialized and fragmented. What makes edge computing valuable to an organization is the operational and logistical efficiencies that can be achieved through real-time analysis. That real-time analysis requires enterprise level server technology that can be designed for edge environments (weatherproof, ruggedized, etc.). Companies like Dell Technologies and HPE are able to design server solutions that meet these very specific industrial environments.
Outlook for the market – is the answer 42?
The edge market is hot and continues to grow at what seems like breakneck speed. Organizations planning edge deployments should fully consider the technical and cultural barriers to adoption and develop comprehensive plans of attack with assistance from trusted partners that have experience.
Who wins and who loses in this market? I believe the market is both big and specialized, which should allow for multiple winners. However, establishing the right strategic partners is a fundamental element to any successful player.
Look for periodic updates on the edge market as we track its growth and some of the emerging players. I’ll leave you with a Douglas Adams quote, from the Hitchhiker’s Guide to the Galaxy.
“A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools.”