Over the past few years, we’ve seen grid and IT technology vendors change their approach to the utility data analytics market. Broad terms like "big data" have dropped out of vogue -- perhaps because utilities aren’t really dealing with data on the global internet scale encompassed by that term.
Data analytics that can cut unnecessary costs and discover untapped revenue opportunities for the people who actually do the work and make the decisions at utilities, on the other hand, can be a valuable commodity. The software that does this tends to go by names like "operational insight" or "situational intelligence" -- or, in the case of Silver Spring Networks’ new platform announced Wednesday, “Operations Optimization.”
The data side of this new platform is Silver Spring’s SilverLink data platform. This started as a cloud-based data collection and integration platform for the 30-million-plus smart meters and other endpoints it has networked for utilities around the world. It has since expanded to actively manage data and tap the computing capabilities in the network itself, and open it to use by utilities and partners.
The analytics smarts, meanwhile, are the “next generation of the flagship solution from Detectent,” the startup acquired by Silver Spring in January. Detectent is mainly known as a provider of analytics to support “revenue assurance,” that is, loss prevention and theft detection. But it’s now powering Silver Spring’s three other new modules -- AMI operations, grid operations, and customer programs -- as well.
These modules include some interesting capabilities on Detectent’s part, including an energy disaggregation technology -- using data to figure out what portion of a building’s load comes from air conditioners, appliances, pumps, and other big electricity users -- that’s being used by flagship Operations Optimization customer Glendale Water & Power.
“Many customers in the industry probably view us as more of a revenue assurance provider,” Mike Madrazo, Detectent’s founder and now vice president of analytics product management for Silver Spring, said in an interview this week. “The other modules are really a very fast-growing part of the business. I think we’re a good way now toward truly integrating the business to accomplish that.”
Detectent got started a decade ago, when most utility billing was monthly, and only larger commercial and industrial customers had interval meters to collect hourly data. It made up for this lack of granularity with data from a variety of non-utility sources, including public-property records, business records, postal records, inventory management systems, and the like, he said.
"Non-technical losses" -- the nice way to describe electricity that’s been unmetered, improperly accounted for, or stolen -- are a well-identified cost for utilities, ranging from low to medium single-digit percentages in most of North America and Europe to up to 30 percent in countries like Brazil and India. That’s made it a focus of a number of utility analytics offerings, with smart meter vendors, utility software providers and numerous startups providing different solutions.
Detectent has built up a sizable list of customers for its version, which includes supporting modules for the field investigators, collections agents and work order management systems involved. But it’s also been putting together a whole new set of analytics for tasks like meter tracking, grid operations and energy data disaggregation, Madrazo said.
“Usually for us, it starts with a partnership with a utility,” he said. “Once there’s a module up and running, and everyone’s using it, I usually dream up the next one.”
Each of Silver Spring’s new modules is building on work with certain utility projects, he noted. The grid operations module, for example, builds on work Detectent had been doing for Chicago utility Commonwealth Edison to help it meet the performance metrics imposed by the Illinois state legislature as part of the deal that allowed the utility’s $2.6 billion smart grid plan to move forward.
Silver Spring is ComEd’s AMI vendor, and Detectent has been integrating Silver Spring’s meter and grid sensor data to help the utility true up its distribution grid models with real-world data, discover phase imbalances, track transformer loading, and other such tasks, he said.
As for energy disaggregation, that’s part of the new “customer programs” module, which is being used by Glendale Water & Power, he said. Last November, the Southern California municipal utility picked the startup’s $280,000 bid over offers from Oracle and C3 Energy priced at $1.2 million and $2.8 million, respectively, largely because the company “had this disaggregation technology, and [the utility] would have had to go and develop it,” Madrazo said.
Detectent’s approach to energy disaggregation differs from those developed by startups like Bidgely and PlotWatt, both Silver Spring partners, he noted. Unlike technologies that use circuit-level or meter data exclusively, “We started creating ours in 2009, when AMI data was hourly,” he said. That’s not the best resolution for accurate disaggregation, so Detectent adds in historical data to build daily load profiles; facility and business records to model how different customer classes use energy in varying ways; and other insights derived from non-utility data.
Another difference is that, while disaggregation products from Bidgely and PlotWatt have been aimed at providing their data to customers, “ours are utility-facing,” he said. Detectent’s disaggregation is primarily meant to serve as a tool for utilities to perform more accurate load forecasts, pick out the best targets for energy-efficiency and demand-response programs, and better measure and improve upon their effectiveness.
Glendale “just went live a few weeks ago -- they’ve purchased the revenue assurance module and the customer module,” he said. “They’re using it for targeted marketing right now,” but “they want to get into M&V [measurement and verification] for programs.” Disaggregation could help them see which customers are good candidates for certain appliance rebates, or pool-pump load control programs, for instance, and then track how much value those programs delivered compared to what they cost.
Silver Spring is far from the only major smart-grid vendor seeking to build analytics platforms and services on top of their established base of meters, sensors and networks. Itron, Landis+Gyr, Sensus and Elster have all been working on different approaches to this common goal, through internal R&D, via partnerships, or by acquiring companies with utility custom in one or another area of analytics specialties.
At the same time, “big data” competitors such as Oracle, IBM, SAS, SAP and C3 all have their place in the utility space, Madrazo said. But they tend to work on enterprise-scale projects, whereas AMI and grid vendors are building on their strengths: millions of smart meters in the field -- and the incentive to optimize the value of each of them.