Despite the hype, the benefits are real -- we just need the analytics to catch up
2012 was the year that “Big Data” went truly mainstream, making the leap from technical concept to marketing buzzword. Although it’s difficult to define precisely what Big Data is, there are no limits to the miraculous things it can do. Big Data can beat you at Jeopardy. It can tell you if your teenage daughter is pregnant. It can even win presidential elections.
Big Data has all sorts of implications for energy efficiency. For example, Nest began life as the iPod of thermostats, a sexy but seemingly niche item pitched at energy-conscious homeowners with a sense of style. More recently, however, it’s become clear that Nest’s ambitions are actually much larger. The networked thermostats are backed by a great deal of computing power that weigh factors such as weather, occupant behavior, and grid load to help utilities design and deploy incentive programs that result in measurable efficiency gains.
On the commerical side, Microsoft recently made a splash with an interesting if somewhat overblown look at its own corporate campus as a template for the “city of the future,” in which thousands of networked sensors gather hundreds of millions of data points on a daily basis, allowing building engineers to detect problems and deploy fixes with unprecedented speed.
The concept of Big Data has taken off at least in part because it’s easy to grasp: as more people and devices get wired up to the internet, every click, call, kilowatt or other formerly analog event leaves a digital residue that can be gathered and mined for insight. Before smart meters, you were lucky to get twelve energy data points per year. Now you get 35,000.
The term Big Data, however, obscures the real story, which is the rise of a data analytics as a critical management tool across a wide spectrum of human endeavors. Of course, data has always been central to any complex undertaking. You can’t manage what you don’t measure, as the cliché goes.
The difference is that, in the past, analysis sought to simplify complex data sets by summarizing and aggregating. Corporations boil down millions of individual transactions into profit and loss statements. Property managers characterize enormous building using simple energy intensity figures.
These aggregate statistics can be useful, even elucidating, but they also tend to sand away many of the bumps and twists that make Big Data so potentially valuable. That energy spike in the middle of the day will never show up in your kWh/sqft, but it could easily drive up your bill by several percentage points. Of course, no one has the time to monitor the torrent of raw data generated by a modern building each day, which is why better analytics have such an important role to play in the future of energy management.
It’s early days for this stuff. We’re still mostly in a world in which “data analysis” consists of alarms based on threshold detection. This hardly counts as analysis at all, but it at least helps pull a weak signal from the noise. The real change will come when the generation of startups that include companies like Gridium succeed in embedding expert building knowledge in their systems, delivering up insights to property managers and engineers that pretty soon they won’t be able to believe they lived without.