Today, Amazon Web Services is holding its annual energy symposium in Houston, where it brings together its customers and partners to discuss the energy sector’s digital transformation and embrace of artificial intelligence.
Like other big tech companies, AWS works with energy providers across the spectrum — from oil & gas to clean electricity. I talked with Howard Gefen, AWS’ general manager for energy and utilities, who gave a preview of how the company is framing this “defining moment” for AI in the power sector.
In part because of projected load growth in the coming years, the energy sector “requires a fundamental transformation in the approach to innovation, from everything from digital operations to emergency response … and at the heart of the transformation is modernization of enterprise systems and infrastructure.”
Gefen explained how AWS is working directly with energy customers — who he said are looking to increase both capacity and efficiency, and ultimately lower costs — to build customized solutions using their proprietary data.
Duke Energy, for example, created specialized applications for power flow analysis that reduced runtime from days to minutes. PG&E developed AI-powered vegetation management systems with AWS that direct maintenance crews to problem areas rather than having them make scheduled rounds across entire regions. And GE Vernova partnered with AWS to develop its grid orchestration software, a suite of tools utilities can use to increase renewables penetration and reduce network outages.
Of course, the pillar for these tools is data: “We’re moving from general models and tools that get used by everybody to really working with companies to create their own models, using their proprietary data that they’ve gathered for years,” explained Gefen.
Gefen has been in his current role at AWS for about five years, and in that time he’s witnessed companies get much more serious about how they gather data. Processes that once took hours of manual reporting, now are incorporating artificial intelligence and other digital tools.
“I’ve seen an evolution from ‘We have a lot of data and we’re not sure what to do with it,’ to ‘We have data, and we need to have practices on how to ingest it, store it, manage it, and use it.’… The vast majority of the companies we’re working with are on a journey to make [data] more accessible and valuable.”
This is not without challenges, as the data is often siloed and unstructured — kept in thousands of spreadsheets, PDFs and databases.
“In large, global companies data is gathered over years and sometimes decades in an area of a business, and it’s used for that area only,” Gefen said. “It’s only in recent years that companies have taken a much more holistic view, and said ‘Oh wait, we need to use all our data together. Different business units need to have access to all our data so that they can make better integrated decisions.’”
Gefen called it a “systematic transformation” that is opening up new possibilities for AI applications across enterprises in the energy sector. The companies that master this shift, he said, will have a decisive advantage in managing the grid of the future.
A version of this story was published in the AI-Energy Nexus newsletter on June 4. Subscribe to get pieces like this — plus expert analysis, original reporting, and curated resources — in your inbox every Wednesday.
