A shift huddle doesn’t sound like the front line of an AI revolution.

These are typically blunt, practical meetings. A supervisor, some engineers, maybe a few maintenance and production leads standing around a whiteboard or a screen discussing any problems encountered during the latest shift. Maybe something went wrong on the line or a conveyor behaved oddly. A supplier shipment arrived late. Any of these will put the production target at risk.

The question at the heart of these huddles used to be simple, and slow to answer: “What happened and how do we fix it?”

At GE Appliances, that’s starting to change.

Now, shop floor and logistics teams can pull up AI systems that surface anomalies in real time—flagging patterns, trends and early signals of failure before they become disruptions. Instead of spending the first half of the shift gathering data, teams are increasingly moving directly into problem-solving.

And behind that evolution is a growing network of AI agents, hundreds of them, not centralized in a data lab but used by the teams close to the work.

A system built for a less predictable world

The backdrop to this change isn’t a single event. It’s a sustained condition: manufacturing and supply chains operating under persistent variability.

If you’ve been manufacturing anything over the last decade, it’s likely you are very familiar with this situation. Inputs fluctuate, supplier timing shifts, constraints tighten and loosen. And production schedules require constant adjustment.

Pour lire la suite : https://www.engineering.com/inside-ge-appliances-ai-agent-rollout/?spMailingID=201567&puid=3218464&E=3218464&utm_source=newsletter&utm_medium=email&utm_campaign=201567