Big data, now and forever
More data, more often, from more sources—that’s the big data mantra. Despite a reputation for being slow-moving with regard to technology, the utility industry is moving to a new level in analytics driven by the explosion of data from smart meters and other sources.
That’s not really news. But what most people never contemplate is what happens to all that data after five, ten or twenty years. The challenge over the longer term is how to manage large volumes of data that are not only growing in size but changing in nature.
Ventyx Senior Director of Product Development Steven Zyglowicz made this distinction in his presentation this week at Ventyx World, and I have to admit I’m among the aforementioned “most people” who hadn’t considered the lifespan of data.
Zyglowicz cited text as one example. He sees it as “semi-structured” data, something we humans can understand without an intermediary but that remains essentially opaque when we’re confronted with vast amounts of it. How do we extract meaning from years of maintenance reports, for example?
That’s where analytics comes in. Zyglowicz painted a tantalizing picture of a future, not that far away, in which our analytical tools not only sift through oceans of codes and numbers but also work in natural language (i.e., the kind we speak to one another).
This kind of flexibility will become increasingly important not only because utilities and other businesses are collecting a wider variety of data types, but because a given data stream may even change as technology advances. Voice-to-text, for example, could very soon replace the typing process for field crews and technicians performing inspections on vital equipment. Just speak into the microphone, sir, and everything will be automatically entered into the database back at the shop. No keystrokes required.
Another shift Zyglowicz noted will be the disappearance of data archives. With ever more powerful computing platforms and cheaper storage, the production environment IS the archive.
Long live big data.