C3 Energy, the well-funded and well-connected startup founded by software billionaire Tom Siebel, has been pretty quiet over the past year about how customers were putting its promise of a new level of data analytics for the smart grid to the test.

No longer. On Wednesday, the Redwood City, Calif.-based startup announced a major new project with Baltimore Gas & Electric, one the utility hopes will save it hundreds of millions of dollars by using analytics to manage operations of the 2 million smart meters it’s deploying, and by tapping their data to discover and prevent energy theft and revenue losses.

BGE parent company Exelon also plans to apply the same C3 platform for projects at its Chicago-based Commonwealth Edison and Philadelphia-based PECO utilities, which are deploying a collective 8 million smart meters between them, Siebel said during a press conference at the DistribuTECH conference in San Antonio, Texas.

For a company that believes its full suite of analytics capabilities can add up to $300 in value per meter per year, that could add up to billions of dollars in value for Exelon, he noted. Although, given that the companies haven’t disclosed what the underlying analytics platform has cost to implement, it’s hard to calculate an ROI.

So far, C3’s data analytics have been focused on the first widely deployed sources of grid edge data, the smart meter. “What we’re doing is aggregating all these data, […] trillions of data points that we’re regressing against each other, correlating against each other, and from that we can extract value,” Siebel said.

In the case of BGE, Christopher Burton, the utility’s vice president of smart grid, said it’s targeting “where we can get the value most quickly. We soon saw that revenue protection, from a top-down analysis, was really the quickest hit.” According to C3’s value calculations, developed from work it did with McKinsey & Company, BGE believes it can better detect, isolate and reduce meter tampering, unbilled energy delivery and other revenue losses to the tune of $6 to $8 per meter, for a $24 million annual impact.

But with the groundwork laid with the initial implementation, “we will look at additional business,” he said. “The service bus already has a lot of data there, so [we] can add new integration to possibly add new modules” from C3’s suite of applications, ranging from new modules for outage analytics and prediction, customer market segmentation, voltage optimization, along with its previously announced apps for asset risk and grid investment planning, smart meter deployment and customer reliability and safety.

Getting all the data C3’s cloud-based data analytics engine puts to use required integration of about 40 different forms of data from systems like meter data management, outage management, customer management, billing platforms and asset management, he said. It also integrates more real-time data from BGE’s Silver Spring Networks Utility IQ smart meter platform, a testament to C3’s ability to manage both streaming data and batch data from a single system.

That’s a pretty hefty amount of data integration to handle. But with the project beginning in October and scheduled for implementation this spring, “I would say this is a very aggressive and accelerated implementation,” he said.

While it’s possible to find lower-cost analytics solutions that pull data from one or a few of the multiple sources C3 is working from, “We went through an extensive PRF process,” he said, and found that C3 has “the architectural platform that we believe has long-term value, not just for today’s applications, but for future applications we hope to deliver.” Beyond that, as ComEd and PECO start to explore how they could put C3’s platform to use, “When they’re ready, it’s much less incremental work” to bring them onto the same platform.

Indeed, with its new analytics modules, C3 now covers a pretty wide swath of the business processes being targeted by many of its data analytics competitors. That’s a long list, including IT giants like Oracle, IBM, SAS, Teradata, EMC and SAP, grid giants including General Electric, Siemens/eMeter, ABB/Ventyx, Schneider Electric/Telvent, Toshiba/Landis+Gyr, and startups like AutoGrid, Trove, Opower and Verdeeco, to name a few. All are competing to bring a set of IT tools and capabilities that are largely new to the utility industry -- and few have laid out ambitions as broad as those set forth by C3.

With more than $105 million in venture capital raised to date, C3 does have a lot more money than its startup competitors. Siebel, who sold the customer relationship management (CRM) software company he co-founded, Siebel Systems, to Oracle for $5.8 billion in 2005, also has friends in high places. C3's board of directors includes former Secretary of State Condoleezza Rice, former Energy Secretary Spencer Abraham, former U.C. Berkeley Dean of Engineering S. Shankar Sastry, and former Constellation Energy CEO Mayo A. Shattuck III. As Siebel put it in a Wednesday interview, “We have unusual access to human capital, and we have unusual access to financial capital.”

At the same time, C3 hasn’t pleased all comers. One early partner, General Electric, appears to have moved away from its relationship in favor of new partners like Pivotal to assist its internally developed analytics platform, for example. And former DOE Assistant Secretary Cathy Zoi left her position as C3’s chief strategy officer in August, prompting industry observers to speculate on broader shakeups inside the typically unforthcoming company.

Indeed, Siebel detailed how C3, founded in 2008, underwent a “major pivot” between 2011 and 2012, when it realized that the original target market for its analytics engine, managing carbon emissions and sustainability for big corporate customers, wasn’t going to emerge. Even so, “the fundamental work that had been done was all applicable” to its new utility and energy focus, and he expects 90 percent of its business going forward to be on the grid side of that market.

With Accenture as its integration partner, C3 is targeting more big utilities in North America and Europe, including one European utility that’s soon to announce a partnership around smart thermostats and automated home energy controls, he said. That’s an extension of C3’s customer-side energy analytics programs, which range from enterprise-scale for customers like Cisco and HP, to small commercial and residential scale with utility customers like ComEd and Pacific Gas & Electric.

On the smart grid side, C3 is working with utilities including Southern California Edison, Northeast Utilities, Entergy and DTE, he said, though he didn’t provide details on those projects. To improve the predictive capabilities of its platform, the company has also enlisted the help of experts in machine learning, including Carnegie Mellon University’s J. Zico Kolter and UC-Berkeley’s Henrik Ohlsson, he said.

As Ed Abbo, C3’s CTO, explained, “You can’t encode every rule for theft patterns, because they change over time. Machine learning is used where you have to learn from the data without being able to code an algorithm,” by “training” the computing platform to recognize patterns that emerge from its regression analyses and creating new algorithms that further refine the process. It’s the same method used by companies like Amazon and Netflix to create individual customer recommendations for books and movies. 

“We’re applying machine learning for all our applications,” he said. “It’s a very key and important areas of differentiation for us.” C3 isn’t the only grid data analytics company to cite machine learning as part of its suite of tools, but it does represent another cutting-edge claim to apply techniques as yet untested in the utility field.

Siebel said he’s aware this all may sound too good to be true for an industry that’s just begun to explore the possibilities of data analytics at this scale. In fact, he recounted a recent conversation he had with Elisabeth Brinton, the former Loudcloud executive and chief customer officer of Sacramento Municipal Utility District, who recently joined C3 as EVP of operations, in which he asked her what the utility industry thought about what C3 was offering.

“She told me, ‘Nobody believes it’s possible,’” he said. “I said, ‘That’s great,’ because it is true, and we can do it.”

Tags: autogrid, baltimore gas & electric, big data, c3 energy, data analytics, emeter, general electric, ibm, investors, oracle, sas, smart grid, smart meters, teradata, vc