Perspectivas del sector, casos de estudio y últimas novedades.
Slow changeovers usually mean tooling, programs, materials, and people are out of sync—not that the plant lacks a full 3D model. Standardize changeover elements first; digital twin can wait and must not replace shop discipline.
Kanban walls fail when numbers don’t match what the shift lead must handle that shift. Pick metrics from their daily work and boards become tools—not décor.
Once work leaves the plant, planning becomes WeChat expediting. The first layer of multi-plant coordination is not advanced APS—it is bringing outsourced work orders, dates, receiving QC, and WIP quantities back to one plan table. Invisible outsourcing is a black hole in planning.
One-item-one-code sounds complete, but SMEs often stall on labeling cost and floor habits. Control batches first: same batch, same process, in/out records—then traceability has a usable grip before serial numbers.
Firefighting looks diligent yet lines stop at peak load. Predictive maintenance isn’t big models first—it’s seeing equipment state, work orders, and spares on one timeline.
Line-side storage serves takt, not hoarding. Gut-feel replenishment adds a box and produces overstock and stockout on the same line. Thresholds, containers, and feedback must match shift usage—lean but unbroken.
Visual inspection is not better earlier or only when rolled out plant-wide. Place it at stations with highest customer complaints, costliest rework, or irreversible defects—and accept by false-alarm and miss rates, not camera specs.
Dropping process-industry batch models on a machining shop fails; dispatching reactors like discrete work orders is just as awkward. Discrete and process need different digital paths—recognize production form before choosing modules.
Equipment is connected and data flows, yet the line still runs on paper work orders. Platform value is not connection—it is turning machine state into work orders, maintenance, and takt decisions. Connected is the start; used is the platform.