If Steve Jobs designed grid-enhancing technologies (GETs), what would they look like?
Heimdall Power chief executive officer Jørgen Festervoll reckons they’d be awfully similar to his company’s Neuron, better known across the power industry as the “magic ball.”
“It looks like something Apple would’ve produced, right?” Festervoll suggested in an interview with Factor This, pointing out the device’s simplicity, ease of installation, and lack of required maintenance.
“When I talk to an investor, I say we’re making the Apple Watch of the power grid.”
Plenty of us wear biotech like the rubberbanded wrist apparel pushed by the famous fruit-logo’d member of the Magnificent Seven, Festervoll surmises, so why wouldn’t we want similar monitoring capabilities strapped to the largest (and arguably most important) machine man has ever built? Think of the power grid as a highway, the slickly dressed CEO suggests.
“It’s limited by temperature; that’s your speed limit. If you don’t have a sensor on it, it’s like driving without your speedometer… We’re just giving you a speedometer so that you can access (the capacity) all the way up safely to the speed limit.”
“Everybody gets that, everybody understands that,” Festervoll reckons.
That’s an easy sales pitch, but it gets even simpler. When the face of Heimdall Power is spreading the gospel of GETs to electric utility executives, he puts it this way: Can you think of any strategy or solution in which you would choose to know less about what’s happening on your lines?
“I don’t think there’s a reasonable argument where you’d rather have less information, not more,” Festervoll concludes. “So we need more of this technology.”
Fortunately for Heimdall Power and numerous electric utilities searching for capacity and/or increased visibility at the grid edge, installing the mass-producible “magic balls” is a pretty quick and painless process, thanks to another sort of sorcery: drone technology.
Droning on and on
In March of last year, Heimdall Power and Great River Energy (GRE), a power cooperative providing electricity to 1.7 million customers in the Upper Midwest, announced the largest Dynamic Line Rating (DLR) initiative in the United States to date, the deployment of 52 Neurons throughout GRE’s grid. The idea was to increase the transmission capacity of existing infrastructure by providing real-time measurements, including line current, angle, and temperature.
Initially, GRE wanted each device to be put up the old-fashioned way, by linemen in a bucket truck. That’s easier said than done, considering some dangerous and hard-to-reach terrain in its territory. After a little back-and-forth, the co-op agreed to allow one drone installation with an audience and cameras present.
“Once they saw it, they were like, can we do all of them with drones?” Festervoll recounted with a chuckle, “And so we did all of them with drones.”
Even Heimdall Power’s CEO couldn’t have predicted how much drone tech would take off.
“A few years ago, I would’ve said the drone thing is going to be marginal,” he admitted. “Five percent of our installations, maybe. Now it’s more like 90%.”
While drones enable faster and more cost-efficient deployments of GETs, for the International Brotherhood of Electrical Workers (IBEW) and other unions, giving up gigs to machines can be a mixed bag. But if their core duty is to protect linemen, Festervoll believes drone tech is one way to do it.
“They see it, and then they’re like, this is protecting the linemen, and then they’re positive to it.”
And if electric utilities are classically criticized for being slow, always putting system reliability ahead of speed of deployment on the ol’ priority list, GETs like the Neuron unlock a new gear for getting stuff done. Heimdall Power started its pilot with GRE in September 2023 with four Neurons. Six months later, the cooperative had results.
“You don’t need to pilot something for three years to get it right,” concluded Festervoll.
Having originally set out to determine and resolve the causes of congestion on one of its key lines, GRE’s collaboration with Heimdall Power resulted in as much as a 42.8% increase in electricity capacity on a single line and expanded capacity by an average of 25% across the pilot area. The co-op decided to buy another 50 sensors.
“We signed the paperwork, we produced the sensors, we shipped them to the United States, and we got them installed in eight days,” boasted Festervoll. “I mean, I can’t even get a meeting with most utilities in eight days.”
It took just two more days for Great River Energy to get the dozens of Neurons to work across their operational technology (OT) systems, which were quickly integrated into the gunkworks of grid operator Midcontinent Independent System Operator (MISO).
“The whole period from the pilot to the start of real, live data going into MISO, affecting operations, took 12 months, and nothing in this industry takes 12 months,” Festervoll asserted. “It just shows that utilities that really want to can make this happen pretty fast.”
Could artificial intelligence (AI) make it happen even faster?
Access to power has become the key driver of AI and data center growth, prompting the need for new solutions among utilities, developers, and other stakeholders. This demand growth will test grid reliability, requiring new ways of collaboration and policy structures.
A new DISTRIBUTECH event, DTECH Data Centers & AI, will help stakeholders in this burgeoning industry navigate power constraints and project delays while balancing the demand for sustainable, flexible solutions. Influential executives and thought leaders will assemble in San José, California from May 27-29, 2025 to discuss how AI is transforming the utility landscape and explore ways power providers can manage increasing demand while utilizing the revolutionary tools promised by new technologies.
Registration for DTECH Data Centers & AI is now open – secure your spot today!
A Variable in the AI-Energy Equation
“AI needs energy, but also, energy needs AI,” offers Festervoll, opining over how the potentially transcendent technology seemingly looms over all. “And I think we’re meeting AI in two ways.”
One way is working with the Metas and Microsofts of the world. When a large load customer comes to a utility seeking to interconnect, it can be a pretty binary transaction. Let’s say Google needs 300 megawatts (MW) for a data center. If that capacity doesn’t exist, options are limited. Either Mr. Google goes somewhere else or decides to start the arduous process of building and interconnecting generation.
As the AI race intensifies, “getting in line” is becoming less and less of a viable solution.
Heimdall’s “magic balls” aren’t magic enough to manifest new transmission, but they can play a pivotal role in stacked solutions for data center customers, helping actualize otherwise untenable development aspirations.
In Festervoll’s experience, utilities are accustomed to “big plans, big lines, and big things,” but a modern problem requires a modernized approach. Instead of waiting for a transmission line to be reconductored, what if a utility and its would-be customer worked together to, let’s say, deploy GETs to help open up capacity for some on-site solar photovoltaics (PV) and batteries (preferably while exploring demand response flexibility)? That can be a recipe that works.
Lengthening gas turbine lead times and the relative infancy of the U.S. nuclear nouveau movement make solar PV and onshore wind power the best available options right now, regardless of how some influential federal figures might feel.
“With wind, we’re like the perfect solution,” explained Festervoll, “Because the wind produced by the windmills helps to cool the line, so you actually have a lot more capacity when you need it.”
The second place Heimdall Power comes into contact with AI is via its own tech. Its end game is not to install Neurons in a hodgepodge fashion on select lines across the grid, but rather something closer to world domination.
“We actually believe that you need these sensors on every line in the grid, not because of capacity, not because of single issues, but to understand power flow and optimize it,” offered Festervoll. “And in that space, AI will have a huge influence on understanding and giving you insight.”
At some point, power flow is going to be more or less automated by AI, Heimdall’s soothsayer predicts.
“AI is going to route power to things faster and better than we’re able to do, but it’s going to take time,” he contends. “You don’t go from driving your car directly into a fully autonomous car. You go into a Tesla where you can sit at the steering wheel and take over if the AI does something wrong or something that you’re not comfortable with. That’s exactly how it’s going to be in the utility space.”
So, how long will that take, and how many Neurons would be required?
Festervoll estimates it would take about 10,000 to properly cover Xcel Energy’s Minnesota grid, as an example. He figures 45 drone teams could finish the job in just nine months.
“Of course, you can just scale up with more drones,” notes Festervoll. “So basically, if the Department of Energy said we’re going to do this, would it be technically feasible to do a lot of the U.S. in one year? Yes, and that’s mind-blowing, right?”
Power line congestion now costs ratepayers more than $20 billion per year, and Heimdall Power’s CEO believes DLR implementation could cut that figure in half. Plus, FERC Order 881 mandates transmission providers use ambient-adjusted ratings by July 12, 2025, which will further encourage adoption of tech like the Neuron.
“Things are really picking up, and (utilities) see that they need this,” notices Festervoll. “They have to do this. It’s physics. They want to build more lines, they don’t really want to put up magic balls. It’s just they can’t not do that, right? Our job as a vendor is to really make it as simple for them to do that as possible.”
“It’s starting to scale,” he adds. “But of course, it’s slow. It’s new for (utilities), and they have to change. I don’t think technology is the problem, but if you have to change the way you do stuff… Like we’re creatures of habit. It’s people, right?”
Shaking the Magic Ball
Heimdall Power’s CEO says his company is engaging in discussions with (more or less) every major U.S. utility. Festervoll’s team recently closed a deal to conduct a Neuron pilot with a major one in the western U.S. this fall, with an intent to expand deployment if all goes well. The GETs company also just had a separate request for proposal approved elsewhere and has ambiguously hinted at a potential new use case driven by field data.
It’s a champagne problem, but all of that means Festervoll is putting some miles on his tires, like flying from his office in Oslo, Norway, to the new U.S. HQ in Charlotte, North Carolina, or to global transmission and distribution events like DTECH in Dallas, Texas, last month. Heimdall’s booth featured a demonstration highlighting how quickly its sensors can be clasped onto power lines, as shown in the video below.
Festervoll kept a Neuron within arm’s reach; the Wilson to his Tom Hanks. I asked him how the whole “magic ball” nickname came about. He guesses it stemmed from a rough translation from Norwegian that Forbes picked up on.
“But I’m the kind of guy that just goes with it,” he smiled. “You know what? It brings a smile to people’s faces, and it’s fun. We need to have a little fun. Everything can’t be too serious… I’m happy with it.”
And if Festervoll could shake that “magic ball” and ask it a question?
“That ball would tell you: I’m gonna have a lot of siblings.”
