DJ Patil knows a little something about big data. Not only is he credited as one of the people to co-coin the term 'data scientist,' he has served the Chief Scientist, Chief Security Officer and Head of Analytics and Data Teams at the LinkedIn Corporation and also co-chaired a major review of U.S. efforts to prevent bioweapons proliferation in Central Asia.
His long resume in some of the world’s biggest tech startups, along with academia and government positions, could make you think that he can’t relate to a room full of utilities representatives and the vendors that serve them.
That couldn’t be further from the truth. While LinkedIn is a very different animal than Pacific Gas & Electric, which hosted The Soft Grid, where Patil was speaking, there are vast similarities for any organization that wants to leverage its reams of data to make better decisions and improve the product they offer to customers. For utilities, that’s keeping the lights on, but in the future, it could be a host of energy services.
For starters, the dashboard for data should be intuitive. He asked the audience when was the last time they had sat through a multi-hour training session on Google or Facebook. Never, obviously. The systems that any company -- including utilities -- employ, should aim for a similar user experience.
To get those systems, the key is to not pick just one vendor, warned Patil. “We treat ourselves like a Formula 1 team,” Patil said of how they put together solutions at LinkedIn. They work with Hadoop, Oracle, MySQL, Project Voldemort, Python, Prefuse and Microstrategy, just to name a few. The combination allows them a toolbox, rather than a single tool, to get the results they want out of data. The problem of picking just one technology is not unique to utilities; it is an issue for most companies. “We’re so dialed in to picking just one technology,” warned Patil, “but you miss other things.”
For utilities, that means working with vendors that understand that there is no one solution, and that to truly take advantage of the data coming off of everything from meters to transformers to weather forecasts, visualization matters. One vendor likely doesn’t have a full solution, but a good one will have a network of partnerships.
“When data looks nice, people pay attention,” said Patil. “If you’re trying to get people to pay attention, make it look good.”
This was no small point. When Patil pulled up one graph showing how many people used the word 'ninja' in their job description on LinkedIn, he stopped to enjoy the details of the graph. The color of the line, a nice slate blue; the thickness and smoothness of the line, “with a slight drop shadow,” he pointed out. “Make the data sexy; make the data fun,” he told the audience. “People don’t like to look at crap.”
For utilities that are evaluating vendors, he essentially told them to not be afraid to make sure that the data -- whether for grid operators or homeowners -- is attractive and user-friendly. One attendee asked Patil about utilities’ aging workforces, adding that many workers have been working in the same (or mostly the same) systems for decades.
Again, he pointed to an iPad or Facebook, which is intuitive enough it cuts across all age brackets and backgrounds. Patil noted that companies like LinkedIn brings together people at high levels across the company to talk about the user products they’re looking to build. “The best way to solve your data problem is to not have a data problem -- make it a user problem,” he said.
For utilities, it was a theme that came up often during The Soft Grid. The silos that most utilities have make it impossible to think holistically about how best to leverage the data to solve pain points within each silo. The most successful utilities have restructured as they think critically about data to make the most of it.
Obviously, utilities can’t compete with the likes of Facebook, LinkedIn or Google when it comes to recruiting data science talent. However, Patil said there is likely a wealth of talent within many IT teams that companies already have, and there is also an opportunity to open up data to others to breed innovation.
In other words, consider a hackathon.
Greentech hackathons already happen in nearly every city. Many of them leverage some energy data, in some cases Green Button data, to create novel apps. Most of the companies and groups that get involved build consumer-facing apps. But the reality is that with the right partnerships, especially with the right universities, utilities could hold hackathons that could show the possibilities for all sorts of data coming out of a utility.
If nothing else, a hackathon will help energize the utility's internal workforce and encourage them to think outside of the box.
“When people have hackathons inside the organization,” said Patil, “it’s just amazing the amount of innovation and ideas that come out of it.”