Capacity lost to changeovers that were never counted

A loom earns only while it runs. Output reports show what was produced; they do not show what the machine could have produced and did not, or why.

Three different gaps

Changeover, breakdown and waiting for material are all downtime and they have nothing else in common. Grouped together as lost time they invite a single unhelpful response. Separated, each points somewhere specific.

Changeovers respond to scheduling

Frequently the largest of the three, and the most reducible — not by working faster but by sequencing similar orders together. That requires knowing what a changeover between two given specifications actually costs in time.

Comparability is the point

Utilisation per machine per shift is where the real variation shows. It is also a sensitive number, so it is worth being explicit that the target is the schedule and the maintenance regime rather than the operator.