A Silicon Valley smart grid startup is gunning to lower the cost of demand response by 90% while increasing efficiency 30%.
Palo Alto, CA-based AutoGrid was founded by Stanford University professor Amit Narayan. Its most recent hire is smart grid pioneer Chris Knudsen. Knudsen, who formerly ran PG&E's Technology Innovation Center, joins as chief technology officer. AutoGrid has already attracted marquee investors including Foundation Capital, Voyager Capital and Stanford University. What's more, it is leading a $4-million grant project funded by DOE and the California Energy Commission to investigate "highly dispatchable and distributed demand response for the integration of distributed generation."
The core mission
Amit, Chris and company are tackling what may be the smart grid's hardest problem – how to manage it as an integrated system, not just a series of adjacent, siloed apps that occasionally swap data.
Analytics is the firm's core differentiator – ultra-fast analysis of ultra-large data sets. In fact, its stated mission is to "provide a new generation of software analytics to enable a sustainable energy infrastructure of the future."
Other companies have also jumped into the grid analytics fray, but AutoGrid is also grappling with the physics of the grid. In this sense, the company is part of the move to blend IT and OT (to blend data from information technology and operations technology). As a first step, the company began by building a grid simulation system. It built that simulator on top of best practices pulled from places such as Pacific Northwest National Laboratory (its free GridLAB-D tool), Lawrence Berkeley National Laboratory, and Columbia University. And from the financial industry, which has decades of experience grappling with Big Data.
Job One: demand response
With its basic "engines" up and running, AutoGrid is turning its attention first to demand response (DR), where it will try to aggregate disparate systems into an integrated whole. DR is challenging today because it lacks a unified view and a unified control of all demand-side resources. (A need Calico Energy is targeting as well with its unified operations center.)
But AutoGrid wants to go past unification to aggregate and optimize as well. It believes it can someday aggregate such things as residential DR, commercial DR and EV charging into a virtual power plant.
AutoGrid is after next-generation, real-time demand response. The kind of DR that can be dispatched in seconds rather than a day ahead as is typical now. This kind of split-second response would permit DR to participate in markets exactly the way generation does today. AutoGrid describes it as "DR 24x7x365." Part of the company's secret sauce is its new forecasting engine, which relies on "machine learning" to create individualized forecasts for customer loads.
(We've covered this gradual evolution of DR extensively here at SGN, including Enbala Power Networks' automated demand response and Honeywell's selection to manage Europe's first automated DR pilot.)
Beyond unification to optimization
AutoGrid claims it can optimize the system as well. It plans to take data from edge devices and from SCADA systems and from its own forecasting engine. And then to crunch that data to make DR dispatch highly scientific and highly efficient.
If AutoGrid or its competitors can truly make that leap, they will fulfill the prediction of FERC Chairman Jon Wellinghoff that "demand response is smart grid's killer app."
Palo Alto, CA-based AutoGrid was founded by Stanford University professor Amit Narayan. Its most recent hire is smart grid pioneer Chris Knudsen. Knudsen, who formerly ran PG&E's Technology Innovation Center, joins as chief technology officer. AutoGrid has already attracted marquee investors including Foundation Capital, Voyager Capital and Stanford University. What's more, it is leading a $4-million grant project funded by DOE and the California Energy Commission to investigate "highly dispatchable and distributed demand response for the integration of distributed generation."
The core mission
Amit, Chris and company are tackling what may be the smart grid's hardest problem – how to manage it as an integrated system, not just a series of adjacent, siloed apps that occasionally swap data.
Analytics is the firm's core differentiator – ultra-fast analysis of ultra-large data sets. In fact, its stated mission is to "provide a new generation of software analytics to enable a sustainable energy infrastructure of the future."
Other companies have also jumped into the grid analytics fray, but AutoGrid is also grappling with the physics of the grid. In this sense, the company is part of the move to blend IT and OT (to blend data from information technology and operations technology). As a first step, the company began by building a grid simulation system. It built that simulator on top of best practices pulled from places such as Pacific Northwest National Laboratory (its free GridLAB-D tool), Lawrence Berkeley National Laboratory, and Columbia University. And from the financial industry, which has decades of experience grappling with Big Data.
Job One: demand response
With its basic "engines" up and running, AutoGrid is turning its attention first to demand response (DR), where it will try to aggregate disparate systems into an integrated whole. DR is challenging today because it lacks a unified view and a unified control of all demand-side resources. (A need Calico Energy is targeting as well with its unified operations center.)
But AutoGrid wants to go past unification to aggregate and optimize as well. It believes it can someday aggregate such things as residential DR, commercial DR and EV charging into a virtual power plant.
AutoGrid is after next-generation, real-time demand response. The kind of DR that can be dispatched in seconds rather than a day ahead as is typical now. This kind of split-second response would permit DR to participate in markets exactly the way generation does today. AutoGrid describes it as "DR 24x7x365." Part of the company's secret sauce is its new forecasting engine, which relies on "machine learning" to create individualized forecasts for customer loads.
(We've covered this gradual evolution of DR extensively here at SGN, including Enbala Power Networks' automated demand response and Honeywell's selection to manage Europe's first automated DR pilot.)
Beyond unification to optimization
AutoGrid claims it can optimize the system as well. It plans to take data from edge devices and from SCADA systems and from its own forecasting engine. And then to crunch that data to make DR dispatch highly scientific and highly efficient.
If AutoGrid or its competitors can truly make that leap, they will fulfill the prediction of FERC Chairman Jon Wellinghoff that "demand response is smart grid's killer app."
Installations of wind farms with less than 20 megawatts of capacity may rise to a record this year if lawmakers expand a federal tax credit.
Two big wind development projects on Appalachian ridges in Bedford and Clearfield counties have been canceled, and fewer new turbines will be spinning across the nation next year due to the possible end of a federal tax credit program that has driven development.
The world's most efficient solar cells, a new vaccine against chicken cholera and recycling car tyres to make steel are among the five winning inventions at the inaugural Australian Collaborative Innovation Awards.