A lot on the line: Creating an intelligent grid through AI-powered smart transmission

We’re living in an age of smart technology, where we can track our sleep from our wrists and do manufacturing quality checks using only sound waves.

However, when it comes to the electricity grid in the United States, there’s a literal gap in useful knowledge: the transmission lines connecting our power supply (and energizing all our smart devices, along with everything else).

Despite nearly $40 billion in spending through 2024 to improve data collection and automation within the U.S. grid (to say nothing of investments elsewhere in the world), what’s going on along our power lines largely remains a mystery. And a costly one at that.

We need a more accurate, reliable view of what’s happening across the entire line — not based on point estimates from a static model of line conditions or the occasional clamped-on sensor, but actual, real-time data from the entire length of the conductor. This could help significantly improve safety, manage costs, increase the line’s capacity to transmit power, and enhance reliability with more precise insights about events that would trigger an outage.

By reducing operational costs and boosting reliability, we can address energy efficiency much more quickly and affordably than through building capacity alone.

Data is only the beginning. We also need better analytics. And not just traditional “if-this-then-that” type algorithms. The flood of data and split-second decisions needed to keep our grid functioning require a new approach to grid management; it’s a wealth of information, but also more than even the most capable engineers or conventional computing can handle.

Today’s AI can help meet these demands. With the right data inputs, AI offers the ability to remember prior events, reason and solve complex problems quickly, and generate automated responses. Collectively, AI can actively put all this data to use, optimizing and automating the entire system with constant synchronicity.

Our partnership on GridVista is just one example of how we make true grid intelligence a reality by directly infusing AI technologies into the smart grid.

Not just a smart grid, an intelligent one

Last month, CTC Global, a leading manufacturer of advanced transmission conductors and power lines, released its GridVista platform. By threading fiber-optic cable into its high-strength carbon fiber composite core, and connecting it to monitoring technology built with technology from Google Cloud and Tapestry, CTC can turn every inch of transmission into a smart sensor. GridVista takes that data and turns it into better decisions.

It’s a high-wire act in the truest sense, offering operators a more detailed view into the bulk power system that, until now, they could only predict or guess at. Now they'll know with near certainty the condition of the lines, as well as how loaded, stressed, or underutilized the transmission system is.

It’s also an example of the kinds of advances that are possible across our infrastructure without breaking the bank. When most utility assets can run for decades and transmission largely traverses the same old rights-of-way, greenfield development can be costly for customers. Instead, embedded technology and AI systems can deliver value from existing systems through limited upgrades and better use of data.

GridVista is one piece of this puzzle, but there’re many more challenges out there in desperate need of solving, many more that could benefit from an infusion of AI. It’s one of the many exciting instances where we’re bringing new energy to AI by bringing AI to energy.

“A grid that can sense its own health”

This new approach — moving beyond basic automation to an intelligent system that's autonomous and responsive — is something we’ve seen take off across other industries already. It's not just IoT but AIoT — an AI of things. We’re evolving from useful but static point sensors and models that had to be manually processed and automated to something with an AI layer that can put that data to use, making decisions faster and more accurately than we could before.

We're already witnessing AI help embed intelligence throughout the physical world — in hospitals, factories, cities, and supply chains. It’s now coming to the power grid in a very meaningful way, too.

Most smart grid investments have centered on improving substations and the distribution system’s data collection (e.g. smart meters, synchrophasors), automation (reclosers, load tap changers, switched capacitors), and improved operations (ADMS, DERMS, GIS, etc.). In some circumstances, utilities have clamped on an after-market sensors to estimate line conditions, but needless to say, most transmission lines don’t have these sensors.

Yet, we can now do so much more with the insights from better data, like the kind GridVista is designed to deliver.

As CTC Global’s CEO J.D. Sitton recently observed: “This awareness allows for a grid that can truly sense its own health in real time and provide unprecedented awareness of conditions on the entire line. Whether that’s real time storm impacts, ice load, wind load, branches on the wire, temperatures on or under the line. The GridVista system truly represents next generation capabilities. ”

It’s not just detecting potential failures, which can lead to critical outages or, worse, wildfires and other dangers. There’s also the fact that more precise temperature data along the length of the conductor can help optimize line throughput. Furthermore, making decisions can happen faster and with more certainty because we’re not relying on systems that lack near real-time insights.

As it is, transmission systems are run conservatively, for good reason, to prevent dangerous operations and ensure reliable, economic operations. If we have better data and smarter analytics, operators can more finely tune grid capacity and operations, providing more benefits for everyone.

When the wire becomes a sensor

Useful intelligence depends on good data. It’s got to be timely, precise, and drawn from the right places. Right now, the grid isn't providing that. The models utilities rely on are working from estimates, not measurements. That gap doesn't just limit operational efficiency; it limits what AI can actually do to help.

GridVista addresses this directly, in a quite novel way. CTC’s ACCC conductors, which already deliver significantly higher conductive capacity, are further enhanced through the embedding of optical fibers into the line. This turns the wire from a passive carrier into a continuous sensor. Finally: strain, temperature, sag, vibration, all measured across the full length of transmission, not estimated from endpoints.

That precision changes what's operationally possible. When you know a conductor is running cool because it’s a breezy day, for example, you can safely push more power through it, a capability called dynamic line rating.

As with most things AI and cloud, the real magic happens when we combine infrastructure, hardware, and software — in this case, the monitoring and automation side of the GridVista system.

Tools like Google Earth and Google DeepMind’s WeatherNext models offer environmental forecasting capabilities, and an important component of GridVista. This intelligent layer feeds real-time wind, temperature, and weather conditions into the system, making dynamic line rating more precise.

CTC also built GridVista using Google Cloud’s Vertex AI platform, to help access and orchestrate a range of other models, such as one Google and our partners have built for predictive maintenance. These tools can help identify stress points along the ACCC lines before an outage triggers, or a fault condition before it sparks a wildfire. CTC is also using BigQuery to store and organize the continuous stream of sensor data that makes it all possible.

As a sign of confidence in the platform, Tapestry — Alphabet’s moonshot project for energy grid innovation — will integrate high-fidelity data from GridVista into its comprehensive, virtualized grid model. By combining these operating insights with Tapestry’s AI-enabled tools, grid operators and project developers can run sophisticated simulations of line conditions. This precision allows for better planning and operating decisions, enabling partners to meet load growth by finding and harnessing the existing grid’s untapped potential

Building smarter, not just bigger

When any system faces a capacity problem, the obvious answer is to add more. More generation, more lines, more infrastructure. For some challenges, that's the right call. For the bulk power system, building smarter and intelligently is at least as valuable as building more. Crucially these days, it’s often faster, cheaper, and lower risk, too.

Replacing existing lines with much higher performing, intelligent ones — rather than acquiring new land, securing new rights-of-way, and starting permitting from scratch — delivers significant capacity gains at a fraction of the cost and time. Capital goes further. And because the work builds on existing infrastructure rather than betting on long-horizon projects, the financial risk drops considerably. In an environment where demand is growing faster than traditional infrastructure cycles can keep up, that matters.