By Monte Rummer
Every IT professional who has been in the industry loves new technology and new capabilities. IT professionals enjoy examining a way that a problem has been solved and find new and even better ways of solving it. But all of us have also been there more times than we would like in a meeting with IT leaders and they make a statement such as “We need to develop a cloud strategy” or, worse, “One of our goals this year is to implement cloud.” As much as the sound of that excites us, we quickly get the feeling that, if the train we are on hasn’t gone off the tracks yet, it definitely will now. We get this feeling because we know we are going to introduce this technology for the wrong reasons, even if we might actually benefit from it.
Cloud is not a goal; it’s at best a concept of rendering services but usually is used as an ambiguous marketing term for what IT organizations have done for many years already. The cloud market has brought us newer and innovative services. However, services is the key word here and is one that I use in almost every article or architectural work I do for companies. Everything should be thought of as a service. Every time we hear the word cloud, we should transpose the word to IaaS, PaaS, SaaS, XaaS, etc. Again, services are not a strategy. IT goals will consume services. And without that understanding, the services will consume us.
In the last few years, the concept of the Internet of Things (IoT) has emerged and has entered the hype cycle. Cloud, as ambiguous as it is, did bring about an evolution in the method of service consumption. The key words here are evolution and method where we have services that we have used for decades and changed the way we use and consume them.
The concept of IoT is much more ambiguous than the term cloud ever was. IoT is nothing more than a grouping of entities which, through a network, we can control and/or collect information from them. This is nothing new. For years we have used ATMs, gas pumps, aircraft, maritime ships, cell phones, satellites, power grids, etc. All of these are examples of large numbers of entities that companies control and collect data from across the Internet. I think what IoT was meant to do was to make the concept more mainstream and provide capabilities and benefits for other industries that are not already using it.
Both cloud and now IoT make enterprise architecture more valuable and important than ever before. EA provides many benefits, but it can be broken down, for this article, to providing services that align with the goals of the business in an efficient and predictable manner.
Virtualization is a great example that every company can relate to. The IT goals could have been to reduce costs, which a subgoal would be to reduce the amount of physical hardware. There are two acceptable ways of reducing hardware: rationalization and virtualization. Then a good architect would ask, what solutions are available that would allow us to virtualize? We would then conclude that we would use a hypervisor, which in turn would lead us to start researching what hypervisor technologies are available that would fit our requirements. The important concept, no matter what EA methodology is used, is the thought process of starting with the business goals and needs and translating those into IT goals. Then we have an abundance of services that we can use to meet those goals, so our job is then to refine the requirements and find the services and solutions that best fit the goals. Too many times the exact opposite happens.
A good example of a common failure of EA is antivirus software within companies. Every company knows it should have an antivirus solution, but does it know why and what goals it maps back to? The IT goal should be, first, to detect a virus that is present, and second, to respond to it. How many companies implement a solution but forget about a requirement of notification that a virus has been found? The result is that we have companies that have viruses in their enterprise, which the AV solution is remediating, yet they are unaware that they have a major issue. They forgot the requirement of situation awareness through some form of notification and departmental response.
IoT is more of a concept or capabilities than something tangible at this point, in relation to EA. But this could be extremely important and, if thought of in EA terms, can provide greater services. In one of my previous articles, I proposed that the most valuable commodity a business owns is its data. Aircraft used by airlines is a great example of the value of data and the concept of IoT. For years, aircraft engine manufacturers have added hundreds of sensors to their products. These, in turn, can communicate to the aircraft systems, providing a vast amount of data. If a fault is detected, the aircraft will send a message, ACARS, to the airline’s maintenance department, which will evaluate the problem and sometimes contact the pilots in flight to help them with an issue and/or have parts waiting for the aircraft before it lands at its destination. All of this data is also used by the airlines to compute the best ways to conserve fuels and even predict failures.
EA architects can use the IoT concepts to help meet business goals or to provide them with new capabilities. We can even apply IoT to everything as a service concept. As we look at everything IT does and provides as a service, and each service is an object with service attributes, we can use the IoT concept to begin to collect and analyze the data associated with those attributes. In a big data concept, we can take all the data and begin to correlate many different points of data and data types to make our services more predictable and thus increase quality.
There are efforts to put standards around the IoT concepts and community. This seemed to work when NIST defined the five tenets of cloud services, which took some of the ambiguity out of the term cloud. The biggest benefit from the NIST standard was to make standard as to what the IT community should consider a cloud service. Even if a service only met four out of five, it was still a positive step toward making sense of the cloud concept and market.
In relation to IoT, this will prove to be much more challenging. Cloud is very specific as compared to IoT as far as industry applications. Communication standards, for example, would be an issue across different industries. As in the aircraft example, aircraft already communicate with the airline departments via an industry-accepted protocol and media, which is radio. It’s highly unlikely that airlines would change, and they would not have an incentive to. And since the airline industry already has a standard, then why would it change?
In the end, as new concepts like IoT are introduced, they are in fact good concepts. Some do not always translate into products, and some may or may not become services. Some, however, can be evolved into EA concepts that can provide great benefit. Again, they are just solutions to problems to be solved. They can provide new capabilities that IT and EA personnel should be able to identify and convey in business terms to the business leaders. As always, know the business requirements and always make them the center of what we do. Understand what the capabilities IT solutions provide and, above all, make sure that when a solution is selected that it’s tied to and meets a business requirement, not the opposite. A&G
MONTE RUMMER has been in the IT industry for more than 25 years. Currently, he is senior engineering advisor for CSRA. Rummer is also working on his PhD in IT concentrating in global IT, EA, and service management theory. He can be reached at firstname.lastname@example.org.