At LET'S GO! 2024, we hosted a panel of industry leaders for a discussion on data, AI, and how the combination is transforming utilities. The session was moderated by Ed Balbis (Guidehouse) and featured a dynamic group that included Mukul Shekhar (PG&E), Laura Fowler (NextEra Energy), and Kevin Kaye (Alabama Power Company). Together, they explored how emerging technologies are redefining the regulatory, operational, and customer-facing aspects of utilities.
You can catch the full session on demand any time, or read on for the highlights.
One recurring theme in the conversation was how startups can drive innovation in utilities. In discussing the adoption of AI, panelists acknowledged the larger utilities (“the corporates,” they called them) are capable of building their own AI solutions. But the panel repeatedly returned to the agility the utilities gain when they partner with an AI-specializing startup instead.
As Kevin Kaye put it, “yes corporates can come up with these (AI) solutions as well. But what we can’t do is add a feature in a week. We don’t roll out something new without it being a complete iteration. That’s the beauty of startups. You can leave them for a week and they come back with a whole new set of product features.”
Rather than attempting to build AI tools in-house, the panel seemed to agree that utilities should partner with technology partners that align with their needs. That way, they save resources, maintain focus, and ensure they’re using tools that are continuously updated and optimized with industry-wide learnings.
A key point from the session was the immense value of open data in creating transparency that improves regulatory and operational outcomes. When utilities and regulators have interfacing systems and shared datasets, it makes gaps easier to identify.
“When we’re aware of a deviation in performance… we can start mitigating that by starting some actionable opportunities,” explained Mukul Shekhar. “And that’s what you want to show your commissions: ‘Yes we know there’s a potential in how we are compared to somebody else, but we’re working towards closing that gap.’”
Benchmarking against industry standards was another focal point. By using shared technologies and unified metrics to compare operational outcomes, utilities can enhance their own performance while contributing to the broader industry's growth. The culture of shared insight generation is proliferating thanks to the interaction of AI and open data.
There was also fascinating discussion in AI's application for fraud detection.
Ed Balbis shared a striking statistic that reveals the immense scale of data produced by smart meters in the U.S.: “There are 154 million smart meters out there in the United States producing 54 petabytes annually. That's the equivalent of 2.7 million HD movies being downloaded. That's how much data that's being generated annually.”
Utilities can use AI to detect anomalies in energy usage patterns, identify unbilled revenues, and even flag potential cases of fraud. For a company like Florida Power & Light, a 1% increase in billed revenues equates to $100 million annually.
AI is also helping utilities consolidate disparate datasets into cohesive insights. Panelists described efforts to create centralized repositories of metrics, enabling decision-makers to tie leading and lagging indicators into a comprehensive performance narrative. This "KPI tree network" can help predict trends and impacts, offering executives the ability to make informed decisions more efficiently.
The automation potential of AI was another highlight. Whether it’s generating custom reports or visualizing data on demand, AI tools are reducing cycle times and equipping teams with real-time insights.
Mukul Shekhar describes the common scenario: "There's a lot of work we do building charts on big data sets. And then we take it to our executive leadership and then they'll say, ‘well, can you just make these slight changes?’ With artificial intelligence, you're going to have this all at your fingertips.”
The panel concluded with a vision for the future of utility regulation. Self-proclaimed “recovering regulator” Ed Balbis advocated moving away from traditional performance-based rate-making to outcome-based regulation that would be assessed against a robust KPI framework custom made for that jurisdiction. Leveraging AI, regulators could programmatically assess providers against an established criteria of metrics they deemed most important from reliability and efficiency to sustainability and emissions metrics. Then savings can go back to the utilities and be passed back to customers.
Data accuracy and stakeholder trust is an important element of this. Laura Fowler underscored the importance of accurate, traceable data, emphasizing how important it is for technologies like HData to validate information with precise citations. Transparency capabilities like these ensure that stakeholders can trust the insights driving regulatory and operational decisions.
The panelists’ diverse perspectives converged on a common theme: Open data and AI are a transformative pair. Whether through better fraud detection, operational efficiencies, or regulatory innovation—with AI and open data—the industry is poised to deliver greater value to customers and stakeholders alike.
As the session wrapped up, attendees were encouraged to continue the conversation. “Pull these folks aside during the breaks today,” urged Ed Balbis. Join in the conversation yourself by visiting HData.com to watch the talks.