'Big Data': not necessarily a tsunami

Insights from the UAI Summit 2012

H. Christine Richards | Feb 15, 2012

Share/Save  

"Big data is not so much a tsunami, but a big river of information, and every year it gets bigger, it never stops," said Dale Skeen, co-founder and chief technology officer at Vitria, at the Utility Analytics Institute Summit 2012.

Indeed. Data and analytics are not just one-time events and, as we're seeing at the summit in Orlando, Florida, much like a river, they will continue to transform and shape the utility industry. Man, I love good analogies. Let's discuss some key issues we covered on the first day of the summit, along with a few other good analogies.

People

Those of you who know me know that I'm a people person. And not necessarily that I just like to chitchat, but rather that I believe people are a critical part of any smart grid or analytics initiative. In our research at the Utility Analytics Institute, for example, we've determined that about 24 percent of utility companies list the lack of necessary skills and staff as a key challenge in their grid analytics initiatives. As one utility noted in our Annual Grid Analytics Report, "there are not a lot of resources with the broad experience to design and deploy the advanced grid management our company is striving for."

And it isn't just about those people who design and deploy analytics initiatives. Analytics users count, too. At the summit I got some support on this notion, not only in various anecdotal discussions, but from work presented by OSIsoft and Microsoft. Some recent research conducted by the two companies showed that training time around analytics tools is one of the biggest analytics challenges for about 45 percent of utility companies. A big question is how can utilities companies create self-service intelligence?

Places for many types of intelligence

We hear a lot of "real-time this" or "predictive that" with the analytics discussion, but real-time and predictive analytics are still a little ways off, and that's okay. As Rob Massoudi, senior vice president of business development and strategic partnerships with Space-Time Insight said, "The goalpost is predictive analytics, and we're working towards that."

In some cases, however, you may not want  real-time or predictive analytics all the time. I enjoyed a great example from one of the presentations that compared operational intelligence, which focuses on analytics velocity, with the business intelligence that is more retrospective. They are complementary technologies, and one can't really replace the other. Let's say you are crossing a busy street, would you want to know how things were an hour ago or a minute ago? Or would you want some operational intelligence about that car coming right around the corner as you get ready to step out into the crosswalk? At the same time, if you were designing that crosswalk, it wouldn't necessarily be helpful to know just what is happening along the road right now. You'd want more business intelligence, and be able to understand what has happened over time, with traffic patterns and such, to be able to design a crosswalk that best addresses the needs of the pedestrian with an eye towards good traffic flow.


Best quote of the day


There are business drivers pushing utilities to invest in analytics, such as performance improvement, cost savings and improved customer service. At the same time, however, new-found data and analytics capabilities are pushing utilities to find new uses for those resources. As Peggy Clippert, manager of customer research with We Energies, cleverly put it, "I have this great new hammer, and I'm going around the company to find all the nails I can hit with it."

That's it for now from Orlando. Kate Rowland, editor-in-chief of Intelligent Utility magazine, will be providing another update tomorrow as we wrap up the Utility Analytics Institute Summit 2012. Thanks for reading and I'll be sure to tell Mickey Mouse you said hi.

H. Christine Richards
Senior analyst
Utility Analytics Institute

www.utilityanalytics.com
crichards@energycentral.com
720-363-6531

Related Topics

Comments

Being able to analyze big

Being able to analyze big data to extract value from it is key. We are learning how important the role of data scientist is in predictive analytics and business intelligence, for solving big data related problems. HPCC Systems has recently released its open source distributed Machine Learning (ML) library and the underlying linear algebra Matrix arithmetic libraries, to assist data scientists and developers in these type of tasks. These algorithms leverage the distributed nature of the HPCC Systems architecture, providing for extreme scalability in large feature sets, particularly with hundreds or thousands of features and millions of entries. The real advantage is that, now, data scientists will be able to do machine learning with full parallelization, on Big Data. By leveraging the power of ECL on the HPCC Systems platform, the need for employing many Java developers on a Big Data problem is avoided, and this reduces the overall cost and delivery time of the projects. Learn more at hpccsystems.com

Better Answers and Questions

Thanks for the post.  One of the areas where analytics does not get much credit, is it helps organization ask better questions.  Analysis does provide answers but it also identifies gaps and inconsistencies.  The gaps help us ask better questions and develop new hypothesis, which ultimately identify underlying insights and information.

http://rubbervines.com