Now, That's the Way to Approach Your SOA Infrastructure Investment!

A short post today... There's so much good stuff happening but I want to try to blog more regularly. I wonder if shorter posts with less flourish will generate more conversation?

In any case, last night I was reading through Seeking Alpha's transcript of Duke Energy's (DUK) earnings call, and came across the following paragraph that struck me:

"We proposed a $100 million plan to install solar panels at up to 850 sites, including homes, schools, stores and factories. To implement this Solar Distributed Generation Program, we are asking for approval to install, to own, operate and dispatch the solar panels. This initiative will help us meet the state's renewable and energy efficiency portfolio standard. Also it will enable us to evaluate the role of distributed generation on our system, and gain additional experience in owning and operating renewable energy resources."

Let's put this in software terms... they're doing a $100M pilot project, involving 850 sites, to gain experience with a technology that they know is in their future. They have set a few clear objectives, though perhaps I'm reading between the lines. Let me share with you how I read these objectives:

1. They'll learn what it takes to get approval for this sort of thing.
2. They'll learn how to negotiate "right of way" for all these different environments (schools, homes, stores, factories).
3. They'll get real-world experience at what it takes to meet their state's renewable and energy efficiency standards - this means they'll be able to participate with the state as a knowledgeable partner, rather than a defensive infrastructure owner! (I know, no defensive infrastructure owners in your SOA. Right.)
4. They'll get real experience about the impact on their current infrastructure.
5. And, they'll gain first hand experience about the costs of owning and operating this new infrastructure.

SOA What? Face it, SOA is in the future of every technology infrastructure. Do you expect your organization to wake up one morning and just have the required experience? Do you think you'll get it right the first time, without ever trying it first? Do you think you can put every objective for SOA out there, and have a grand plan and just go? That reminds me of a customer I worked with where their RFP read something like... "Your product should support [insert summary of every WS standard], and allow us to [insert every marketing tag line from registry, infrastructure and SOA management company]." That sort of plan doesn't show any insight because it's not based upon experience. Just the opposite, it made me wary because I immediately knew there was a lack of experience.

Well, Duke's objectives sound like good ones for any pilot SOA project. In fact, with some minor rewording, you can use the list above on your own project.

Though, I'm not foolish enough to think anyone will invest $100M in a pilot SOA infrastructure, clearly some pilot investment is justifiable considering the valuable knowledge gained in the process and the long term impact SOA will have on IT infrastructures in every organization.