Plant-floor AI capability build

Context

A manufacturer building AI capability for the plant floor. The visible gap is ML, but the binding gaps are industrial data engineering (OT/IT convergence), edge MLOps, and OT security — the skills that get a model off the laptop and onto the line safely.

The decision

Staff the deployment skills, not the modeling skills: OT/IT data engineering, edge MLOps, and OT security are the plant-floor bottleneck, and floor knowledge must be built because it can't be hired off the shelf.

What most miss

Teams over-index on data science and under-index on the OT/edge/security skills that decide whether AI survives contact with the plant. On the floor, deployment capability is the scarce one.

Stakes

Skip the OT and edge skills and every model works in the lab and dies on the line.