Utilidata announces development of a custom module for AI at edge of grid - The EE

Utilidata announces development of a custom module for AI at edge of grid

Utilidata has announced that the company is working with NVIDIA to develop and deploy a lower-cost, lower-power module to deliver real-time AI at the edge of the grid.

The purpose-built module will leverage the NVIDIA Jetson platform and power Utilidata’s smart grid chip, a distributed AI platform installed alongside electric metres to enhance grid resiliency, lower grid operations costs, and integrate distributed energy resources (DERs) including solar, energy storage, and electric vehicles (EVs) accelerating the transition to a decarbonised grid. The module will be deployed with Utilidata’s smart grid chips beginning in 2024.

“With edge AI, electric grids can be made more intelligent, reliable, secure, sustainable, and efficient,” says Deepu Talla, vice president of embedded and edge computing at NVIDIA. “Over 1 million developers and 6,000 partners and customers have adopted NVIDIA Jetson to benefit many of the industries, and this initiative with Utilidata will bring transformative technology to accelerate the clean energy transition.”

The new purpose-built module will bring the necessary computational power via a graphics processing unit (GPU) to the edge of the grid at a power consumption, price point, and security standard that lowers the overall cost of decarbonisation while accelerating the clean energy transition and improving reliability and resiliency of the grid.

“Utilidata’s smart grid chip, with the first-ever purpose-built AI module for the grid, will transform how utilities operate the grid edge to better serve their customers,” says Josh Brumberger, Utilidata’s CEO. “By working with NVIDIA on a custom module, we are able to make edge AI accessible to all utilities and significantly drive down the cost of electrifying transportation, add more solar and energy storage to customers’ homes and businesses, and greatly increase the reliability of the grid.”

Utilidata began collaborating with NVIDIA in 2021 to develop the smart grid chip using the NVIDIA Jetson Nano system-on-module. These first-generation devices were installed in 2022 for a demonstration project in Lake Placid, New York, at customer locations along the circuit that serves the Olympic Bobsled Complex. Smart grid chips deployed in 2023 with multiple utility partners will be powered by the NVIDIA Jetson Orin platform.

The new purpose-built module contains a specific-sized memory block for grid edge data, optimised clock speeds of the central processing unit (CPU) and GPU to maximise energy efficiency, and a customised number of cores to optimise the cost of the module. It is specifically designed to process measurements like waveform data from the grid and continuously generate AI models for uses such as granular demand forecasting, identifying pre-outage conditions, and interconnecting new DERs to the grid.

This smaller-sized module will be easier to add directly into grid edge devices like EV chargers, solar inverters, and meters. The open hardware and software architecture empowers third-party developers to integrate proprietary software and build new grid-edge applications, using standard metrology, data processing, and DER interaction tools.

The edge of the electric grid, where customers’ homes and businesses connect to the utility system, is becoming exponentially more complex. As more people shift to producing and storing their own energy with solar, batteries, and EVs computation is needed to process massive amounts of data within seconds and provide grid operators with visibility, information, and tools to enhance reliability, enable demand flexibility, and adapt to a new way of operating the grid. Edge AI accelerates complex computation and decisions where data is being generated by energy consumers, rather than sending large amounts of complex data to a central cloud computing data centre, which increases costs and latency, and delays real-time decision making.

Utilidata’s smart grid chip collects and analyses grid edge data to provide utility companies with real-time visibility into grid performance and capacity. This enables utilities to seamlessly add more clean energy and reduce the cost of existing grid operations or upgrading existing grid infrastructure. The smart grid chip processes waveform data with continuous monitoring and pattern recognition to inform real-time decisions on the grid in split seconds, instead of minutes or hours, which increases reliability and resolves issues with power distribution before outages occur.

NVIDIA Jetson is an AI-at-the-edge computing platform with over a million developers using it for edge AI and robotics. With pre-trained AI models, developer SDKs and support for cloud-native technologies across the full Jetson lineup, manufacturers of intelligent machines and AI developers can build and deploy high-quality, software-defined features on embedded and edge devices targeting robotics, AIoT, energy, smart cities, healthcare, industrial applications, and more.

Follow us and Comment on Twitter @TheEE_io

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.