This was a question asked to me this week by an engineer at the New Zealand User Conference, and I stepped back surprised at the question from the remote south island of New Zealand, but pleased. The question was asked with full reality especially around the utility industry such as water etc why would people put in servers today and have all that sustaining hassles, versus just renting space and capability at a managed data center. The point is extremely valid as the networks become more robust, the acceptance of virtualization as the norm to deploy applications, why have the hassle with local servers.
This especially applies to plants with limited IT capability, as a server requires upgrades, virus protection, and OS support, plus it must physically maintained. Also within 2 - 3 years, a server has to replaced either the software has advanced beyond the PC capability, or parts are not available. When thinking of the South Island of New Zealand the picture of remote, long distances on windy roads, by installing a PC a maintenance issue is created. If a basic local capability that interfaces with devices such as valves and instruments, maybe a local driver and basic SCADA, on a “shoe box” industrially suited box with limited moving parts, e.g.| fans and hard drives. All the data is up to the “cloud” or in this case a remote data center where the heavy lifting of the historian, and the execution of supervisory control is performed. Many of you are probably saying not viable for reliability reasons but is that truly the case with the improving infrastructure? BHP and Rio Tinto the two biggest mining companies in the world have both gone live with remote operational centers, where large equipment are operated from 1500km away, yes with local supporting teams. Yes, they have local control available if needed, but in the last 3 years Rio Tinto has not had to use local control except in controlled situations for maintenance etc. So why not do this for distributed water utilities etc, makes a lot of sense?
In another presentation at the conference, for a distributed wind turbine application the company was showing how they were using the “cloud” for connection security to a remote data center for remote data historians, and SCADA. At ARC conference 2 weeks ago in the US there was company showing Gas pipeline control, and leak detection in the cloud. There are 10s of more applications going in, solving the core issue of sustaining remote servers in harsh environments by shifting that computing to a managed environment.
What surprised me when I posted a blog on the cloud about 6 months ago it was one of the least read posts, during last 6 months the acceptance of remote computing capability is growing. In 6 months to a year it will be main stream, and like this engineer down in the south island of New Zealand, he will be implementing architectures with remote computing and a secure cloud as natural.
Certainly through this year Invensys will release managed services to provide the opportunity to move to this architectural freedom, I expect that this will only accelerate into 2014.
So now if you look forward 2020 the networks either wired or wireless will be extremely stable and capable, in remote areas people will use wireless infrastructure naturally, employing cloud capabilities as Microsoft’s Azure for secure managed access, and leasing computing in the ‘cloud” or managed data centers. Even if, there are local controls for high availability, and reassurance, the normal days activities will be using this remote capability, where someone provides a service to maintain, and upgrade the system hardware and sustain a high level of availability.It was refreshing to see engineers embracing this capability, not as a threat but as an advantage that will help satisfy their customers, and applications.