When the internet of things comes to fruition, the economics of big data analytics will start to pay off, said Gaurav Dhillon, founder and CEO of SnapLogic. It will be a world in which sensored devices will grow far beyond the NEST thermostats and Fitbit devices some of us now have.
The foundation of the current data warehouse business was the simple bar code, used to replenish inventory. Data warehouses are now a $30 billion business, he said. Now imagine when there’s input from all these other devices and machines generating and transmitting data. That’s when the “numbers of big data start to ring true,” Dhillon said at GigaOM’s Structure Data conference on Wednesday.
There are already apps that ping a retailer if they notice a shopper wandering aimlessly around a superstore. The app will will send a sales associate to help the wayward shopper find what she’s looking for. “That could be quite useful,” Dhillon said.
Big data in industrial applications is still not being tapped to its full potential, he noted.
There are huge industrial boilers that incorporate a sensor that watches their temperature, he said. “These things cost hundreds of millions of dollars and this sensor looks at the temperature in the infrared spectrum and today just provides a cable feed back to some guy in a baseball cap who watches it. It’s 2013, why couldn’t we put analytics into it that will tell us it will blow in a w eek because we see these hotspots.”
The same sort of technologies could be applied to aircraft engines, cars and other machinery, he said.
Applying the right analytics to the right data, and also ensuring that machine learning data can interoperate with higher level data from CRM or other applications, is where the real payoff of big data begins, said Dhillon, who is an authority on the niceties of data integration. He is former CEO of Informatica(s infa).
Check out the rest of our Structure:Data 2013 coverage here, and a video embed of the session follows below:
A transcription of the video follows on the next page