Two weeks of grid preparation, reduced to hours.

The human problem

New generation, storage, and electricity demand can wait in long interconnection queues while engineers prepare and check complex studies.

The prior constraint

Data preparation spans grid models, physics simulators, scripts, standards, and repeated manual handoffs.

AI’s actual role

Agents orchestrated existing engineering tools and data preparation; engineers retained final decisions.

The documented result

Duke reports reducing preparation from two weeks to hours. LBNL separately reports more than 2,000 GW of generation and storage seeking interconnection at the end of 2025.

Why it may matter

The useful claim is narrower than autonomous grid planning: less preparation could leave engineers more time for review and more studies.

Limitations

The time reduction is partner-reported and not independently audited.

Faster preparation does not establish faster approvals, lower bills, or better reliability.

The system coordinates physics tools; it does not replace engineering sign-off.

Unresolved questions

How often do engineers correct agent-prepared inputs?

Does the workflow reduce total study duration across more utilities?

What audit trail is retained for each recommendation?