Utilities don’t really know what’s happening on the low-voltage distribution circuits that make up the last mile of the power grid -- and putting sensors, smart meters and other devices out there to find out is expensive.

But the growth of rooftop solar PV, plug-in EVs, energy storage, demand response and other distributed grid-edge systems is forcing utilities to find ways to get insight into the distribution grid. Software that can create accurate models and forecasts of the ebbs and flows of power on this grid edge could really help cut those costs -- not to mention play a role in planning for a distributed energy future.

That’s a short description of what GRIDiant, the startup bought by Toshiba’s Landis+Gyr earlier this year, says its software can do. Since its 2000 founding, GRIDiant (formerly Optimal Technologies) has racked up a significant list of big utilities putting its claims to the test, primarily through pilot projects.

Now one big customer is taking GRIDiant’s technology to system-wide scale. Last week, Pepco Holdings announced it's deploying Landis+Gyr’s grid analytics solution across the service territories of its three mid-Atlantic utilities, Pepco, Delmarva Power and Atlantic City Electric. That's a pretty fast turnaround for a project of this scope, considering that Pepco just started piloting GRIDiant’s GRIDplan Reliability platform in January.

The three utilities will “deploy six different analytics applications for visualization, planning and real-time operational analytics support,” according to last week’s release. Those include a broad swath of applications for GRIDiant’s physics-based distribution grid modeling capability, including outage management and voltage optimization, asset management and revenue protection, and load planning and demand response management.

This is a similar menu of use cases being pitched by all manner of big-data analytics players in the grid space, from established giants like Oracle and IBM, to smart meter contenders like Itron and Silver Spring Networks, to startups like C3 Energy and Integral Analytics -- and that’s naming only a few.

But Pepco’s announcement represents the “first major deployment of an enterprise-wide, physics-based grid analytics platform by a utility in North America," Prasanna Venkatesan, executive vice president for Landis+Gyr Americas region, said in last week’s release.

The key term here is “physics-based,” which alludes to the heavy math which GRIDiant says makes its modeling more flexible and accurate than the model-based distribution management systems (DMS) of today, which essentially use technology scaled down from transmission grid modeling. GRIDiant also combines its physics-based models with statistical analysis common to other analytics platforms, which can help find relationships between different parameters where physical modeling can’t, Kelly Dietz, GRIDiant’s vice president of sales, told me in a January interview.

As Andrew Colman, former CEO of the Los Altos, Calif.-based startup, put it in a 2013 interview, “There are people who do power flows, people who do models, and people who do analytics. But we’ve kind of tied them all together, both for planning and operations speed.” A lot of work goes into cleaning up the geographic information systems and asset management systems that serve as the databases of record for all the wires, cables, transformers and other commodity stuff that makes up the distribution grid, since records on repairs and renovations aren’t always kept.

But the end result is a real-time power flow model of thousands of miles of low-voltage grid circuits, able to adapt to changes coming in and update the other utility systems it’s communing with, Dietz told me. Pepco started with GRIDplan, which feeds grid modeling data into how utilities plan for investments in grid infrastructure -- something that states like California, New York and Hawaii are requiring their utilities to do for future years. Other use cases, including real-time operations like conservation voltage reduction, were on the roadmap with Pepco, she said.

GRIDiant has a lot of home-state customers, including Southern California Edison, Silicon Valley Power, Pacific Gas & Electric, Sacramento Municipal Utility District and the California Energy Commission. It’s also been working with Pepco’s neighbor utility, Baltimore Gas & Electric, since last year, although BGE’s decision to hire C3 for its grand data analytics rollout might mean fewer opportunities for Landis+Gyr’s “comprehensive grid analytics solution."

Landis+Gyr’s vision for GRIDiant includes integrating it with its smart metering and meter data management software offerings, as the “next stage of enhancing and optimizing the data platform, so that our customers can have what we call a situational awareness of the grid,” Venkatesan told me in a July interview. It’s worth noting that Pepco is using Landis+Gyr’s smart meters for its mid-Atlantic utilities, although Silver Spring Networks is running the communications, networking and head-end software for Pepco’s AMI system.