Veritone uses AI to improve smart grid efficiency

  • October 14, 2020
  • Steve Rogerson

California-based Veritone is improving the efficiency of smart grids with real-time predictive artificial intelligence (AI).

Its patented offerings support the green energy boom by empowering utilities to predict optimal energy supply mix and pricing to meet grid demand and ensure grid reliability.

Veritone created the AIWare operating system for artificial intelligence. It has entered the energy sector with Veritone Energy, a suite of proprietary, predictive AI offerings to help utilities increase profitability and improve grid reliability as they make the transition to renewables.

“We are thrilled to be adding a new vertical to our growing customer bases in the media and entertainment and government, legal and compliance markets,” said Chad Steelberg, CEO of Veritone. “Every year, over $750bn is invested in global electricity generation and distribution projects, giving Veritone Energy a significant opportunity to showcase how our cutting-edge, predictive AI technology can help the world reduce its dependence on fossil fuels.”

Veritone has already begun to deploy its AI energy technologies with a US utility and is in talks with other potential private and public sector customers and partners.

Veritone Energy’s patented products use AI to optimise energy distribution across diverse energy grids. By applying models, rules and learning to weather forecast, energy demand, pricing and device data, they can help utilities automatically predict optimal energy supply mix and pricing to meet grid demand, in real time. A utility can now more accurately predict how much energy, at what price, to send to which device, and when.

Veritone estimates this real-time forecasting capability will translate to millions of dollars of cost savings for utilities in saved energy, device longevity and energy prices.

The company’s announcement comes as the increasing use of clean energy sources is creating reliability and safety problems for utilities due to the aging US electric grid, which still relies on human decision-making.

Two-way power flow between legacy and renewable devices can cause network congestion, inconsistent energy delivery and blackouts. The need for synchronisation and optimisation so utilities can add renewable sources has never been greater.

“The green energy boom is impossible without digital innovation,” Steelberg said. “Hamstrung by legacy infrastructure, and challenged by electrical grids that are becoming more distributed and complex, the energy sector urgently needs AI to make clean energy production more predictable, reliable and cost effective.”

Weather and energy demand are constantly changing, creating significant challenges for utility companies and equipment providers that need to deliver energy from a wide variety of old and new energy sources consistently and profitably. Day-ahead forecasting of always-changing weather is not enough.

“Unlike other predictive energy providers that use the past to predict the future, Veritone’s patented CDI technology predicts the future using the present,” said Veritone chief energy scientist Wolf Kohn. “Veritone combines current data from multiple sources – weather sensors, load demand systems and distributed grid devices – to predict optimal energy dispatch, allowing utilities to accurately and dynamically meet future grid demand, whether minutes, hours or days ahead.”

He said the company’s active device synchronisation technology ensured safe and reliable operation of distributed energy resources with the grid on a constant basis, rather than relying on historical feedback loops that may not accurately reflect the present state.

“We do this with patented, Hamiltonian-based models and machine learning that occurs both centrally and at the device level,” said Kohn. “This approach enables distributed, autonomous edge device decisions using low-cost processors that can be more cost effectively deployed on-premise, via internet of things devices or in the cloud.”

With a proprietary operating system for AI and more than 20 issued and pending patents covering its AIWare-based energy technology, Veritone’s customisable offerings can be applied to solve a wide range of problems facing the industry, including solar smoothing, demand response, micro-grid synchronisation, intelligent device control, voltage optimisation, regulatory compliance, dispatch optimisation and high-speed energy arbitrage.

Veritone Energy’s flagship offerings include:

  • Forecaster to detect and predict energy supply, demand and price accurately.
  • Optimiser to make AI-based energy supply determinations.
  • Controller for predictive device control and active synchronisation, combining energy sources to satisfy demand optimally.
  • Arbitrage for buying, selling and dispatching energy.

“Veritone’s predictive AI technology brings much-needed optimisation and efficiency to the energy grid, helping reduce energy consumption and provide greater stability,” said John Zangardi, president of Redhorse, a US government services provider. “Veritone’s ability to leverage massive amounts of historical and real-time data to effectively balance energy supply and demand, with automated synchronisation of grid assets for maximum reliability, is a real game changer in the energy industry.”

Veritone is headquartered in Costa Mesa, California, and has offices in Denver, London, New York and San Diego.