Data analytics for smarter grids

MIT paper offers approach to analytics thinking

Phil Carson | Jan 17, 2012

Share/Save  

I had a great conversation with a data analytics practitioner in the utility industry the other day and will deliver a column on that shortly. Meanwhile, I'll use today's column to bring you a few concepts in data analytics that I think will stimulate your thinking.

The key with data analytics, as with cyber security, is the discussion around "how to think about" the subject, not the nitty gritty "how to" do it. 

The article, "Big Data, Analytics and the Path from Insights to Value," by Steve LaValle et al., was first published in the MIT Sloan Management Review (Winter 2011) and is a joint effort by LaValle, global strategy leader for IBM's Business Analytics and Optimization service, and co-authors at IBM and the Review. The report is serialized in blogs, so you absorb its findings in digestible portions. (The study surveyed 3,000 executives in 30 industries across 100 countries and leavened the results by consulting academics and subject matter experts.)

"Organizations that effectively adopt data-driven management are likely to become leading performers in their industries," according to LaValle et al.

Note the catch: "effectively adopt." Therein lies the rub. So, what does effective adoption look like, what are the challenges and steps?

Turns out you need the right people, processes and tools. People who understand how to set up a strategic, enterprise-wide plan, who can provide centralized oversight and who can assist in identifying the single more daunting problem to which analytics can be applied.

In a neat parallel with grid modernization, a utility must identify the business challenge and determine what questions to ask of associated data in order to begin the process. In grid modernization, the mantra has been to determine the forward-looking business case for the enterprise and create a technology roadmap to enable that business case.

The paper makes the case for executive buy-in and cultural transformation within the organization and it suggests a way to accomplish these changes.

"For analytics-driven insights to be consumed—that is, to trigger new actions across the organization—they must be closely linked to business strategy, easy for end users to understand and embedded into organizational processes in order to take action at the right time," the authors stated.

The two biggest hurdles: lack of understanding on how to use analytics to improve the business and lack of "management bandwidth."

To the report's five critical recommendations:

  • Focus on the single biggest opportunity first ("one big important problem") to demonstrate value to one's organization.
  • Begin with questions, not data. Understand and articulate the problem before approaching data to solve it.
  • Make information "come to life," which means articulating use cases so that non-experts get it. That can mean data visualization.
  • Maintain legacy data practices while overlaying new ones. As centralized analytics oversight grows, keep distributed, localized capabilities in place.

 

The paper introduces the notion of "starting in the middle." In other words, grabbing the bull by the horns can produce swift results. It's possible that imperfect data can yield actionable insights that save money or transform business processes. (More on that angle in a moment.) Spending inordinate time obtaining all pertinent data or striving for perfect data will threaten the data analytics mission. Missing data often can be effectively constructed by proxy or extrapolation in order to produce a key insight.

To provide a sense of where all this is going, the authors identify three stages of analytics prowess: aspirational, experienced and transformed. The first group, presumably where the bulk of electric utilities reside today, focuses on efficiencies and automation to manage costs. The second group has moved beyond cost management to optimize the organization and how it accomplishes its mission. The third group uses analytics as a competitive differentiator and has the people, the processes and tools in place to focus on their customers' profitability. This last group is free to make targeted investments in niche analytics applications.

Need an example to get your arms around these eye-glazing bromides? I'll provide one ASAP.

Meanwhile, I'd suggest you explore what's on offer at Energy Central's Utility Analytics Institute's Summit 2012, slated for Feb. 15-16 in Orlando, Fl. And you can peruse a few recent columns on the topic.

"Harnessing 'Big Data' for Business Value"

"Utility Analytics Offer Value Now"

Phil Carson
Editor-in-chief
Intelligent Utility Daily
pcarson@energycentral.com
303-228-4757

Related Topics

Comments

Great Post

The nail was hit.  It is about asking the right questions, which drive how the systems are designed, configured, and implemented.  Asking the right question requires people who have already been through this process.  Going to conferences and talking to colleagues at other utilities, I have not seen too many people who have been through the process. 

Experienced practitioners, speak to me!

Thanks for reading and commenting.

We'd love to hear from experienced practicitioners about their approaches. Just give me a ring or a ping.

Regards, Phil Carson

Hitting The Nail On The Head

This article hits the nail on the head and is a must for mainstream utilities to understand in order to move forward with efficient smart grid implementations.  From past experience in utility operations and many different management opportunities, I would like to add my success experiences to the critical points in the article.

 

The report's five critical recommendations:

  •  Focus on the single biggest opportunity first ("one big important problem") to demonstrate value to one's organization.

·         Focusing on the biggest opportunity gets buy-in across the organization and facilitates cross organizational communications to explore and develop a greater understanding of how it affects many, if not all, of the different operational and financial parts of the business.

 

  • Begin with questions, not data. Understand and articulate the problem before approaching data to solve it.

·         Cross organizational communications becomes the catalyst that drives the process in asking the right questions to get at the best solutions.

 

  • Make information "come to life," which means articulating use cases so that non-experts get it. That can mean data visualization.

·         Once the cross organizational communications has begun and continues, it leads to a greater understanding of who needs what and how they need to see it.  You now become more efficient in moving the raw data from the servers to information formats that shows up on the correct desk and in a way they can understand to act on it.

 

  • Maintain legacy data practices while overlaying new ones. As centralized analytics oversight grows, keep distributed, localized capabilities in place.

·         The new processes have to be understood and then integrated into the legacy process and once tested and understood, then new process handling methods can be developed which efficiently integrates all processes and the new technologies.

 

You need: "the right people"," good processes", and "the right tools" to make it happen.

 

Richard G. Pate

Pate & Associates, Principal

rgpate@pateassociates.com

www.pateassociates.com

 

Follow us on Twitter: @pateassociates

Connect with Us on LinkedIn: Richard G. Pate

Check out our blog for all the latest news: pateassociates.wordpress.com