)
You Cannot Solve What You Cannot See: Physical AI and the High-Resolution Grid
The grid's hardest emerging problems are high-frequency phenomena that the industry has largely been unable to see. Inverter-based resource instability, the volatile load of hyperscale data centers, incipient equipment faults, and wildfire initiation all unfold at speeds and resolutions that traditional SCADA and meter data simply do not capture. You cannot solve what you cannot see, and for decades the most consequential behavior on the grid has been invisible by default.
This session makes the case for Physical AI: the idea that high-resolution sensor data is not a niche diagnostic tool but the foundational layer of the modern grid, the substrate that advanced analytics and AI actually require to be trustworthy. Just as the industry once decided that a data historian was essential infrastructure, the next era depends on treating high-frequency sensor data as a first-class grid citizen, captured at its native resolution before it is lost forever.
Dominion Energy Virginia and PingThings will share what becomes visible, and therefore solvable, once the grid is observed at full resolution. Drawing on a six-year collaboration and real operational use cases, the session shows how seeing the grid differently changes what utilities can ask of it: catching instability before it cascades, understanding new loads as they actually behave, and turning raw measurement into operational intelligence.
The session closes with the larger vision behind this work. Capturing the high-resolution grid is not a single project but a long-term infrastructure commitment, the same kind of commitment the industry once made to the data historian, and the one the next era now requires. The presenters will describe what that commitment looks like in practice: building enduring capability rather than one-off pilots, making the grid fully observable, and ensuring the industry is no longer flying blind through its most consequential moments.
