Delivery reliability begins with measured work content and flow

The standard minute value, or SMV, is a work-content reference with conditions attached. The ILO defines it through the output that qualified workers can achieve without overexertion when they know and follow a specified method and are motivated to apply themselves.[S03] An SMV is therefore not the raw average of whatever happened on a shift. It is not transferable between different operation sequences, machine settings, material behaviours or quality requirements merely because the garment name is unchanged.

Industrial engineering should start with an operation breakdown that describes the work in executable terms. Each operation needs a sequence number, a clear start and finish, the machine or workstation, the attachment or folder, the material and trim condition, the quality criterion and the disposition of the completed bundle. The method release should also identify the style, product family, routing revision and effective date. Without those attributes, a time value cannot be audited against the work that was actually performed.

The observed time and the released standard serve different purposes. Observations describe a particular worker, method and period. The released SMV provides the planning reference for a defined method and qualified performance. The ILO discussion also distinguishes allowances within the calculation, including personal time and machinery-related allowances, before the aggregate SMV is formed.[S03] GFT would retain the observed samples, rating basis and allowance rationale beside the result rather than leaving only a final number in a routing table.

A provisional standard is legitimate during ramp, but it must be labelled as such. Its status should state the evidence still missing, such as a stable method, enough qualified observations or confirmation of the allowance basis. The planner can use that provisional value with an explicit uncertainty allowance. The reviewer should not present the resulting capacity as settled. Once the method changes, the old value remains attached to its historical orders and a new revision becomes effective for future comparisons.

Method control also protects workers. A standard developed for one sequence should not become a personal output target for a worker who has been given a different machine, an incomplete bundle or an unapproved material substitution. Before a variance is attributed to performance, the supervisor should be able to confirm that the released method was available, the operator was qualified for it and the required inputs were present. A work-content measure is credible only when those conditions are visible.

Finally, the SMV does not measure every form of time. It does not replace records of waiting for material, machine stoppage, changeover, absence, rework or dispatch queues. Those events belong in their own logs. Keeping them separate prevents a planner from hiding an operating loss inside an inflated standard and prevents an engineer from treating a necessary allowance as an unexplained delay.

A feasible plan can still become an unbalanced line. The ILO garment-sector report assigns line balancing to supervisors systematically before production through planning and daily during production. It also describes regular in-production checks using recorded output and work in progress to identify bottlenecks and their causes.[S04] The important point is the management rhythm: balancing is not a one-time industrial-engineering calculation handed down with the style file.

The pre-production balance begins with the released operation bulletin, the expected order quantity, the available skills and the machine layout. Work is allocated so that each station has a coherent assignment and the planned load is feasible at the constraint. The supervisor should also know which operations are method-sensitive, which workers can cover them and where bundles will be counted. A balance that depends on an unnamed person arriving with an unrecorded skill is not yet an executable plan.

During production, the line board should show planned and actual good output by interval, WIP at agreed boundaries, operation-level stoppages, quality interruptions and skill absences. The purpose is to see a forming queue before the end-of-line total falls. A station can meet its own local target while starving the next station or building an excessive buffer. The review therefore concerns the whole sequence, not isolated operator utilisation.

A rebalance should follow diagnosis. If the cause is a wrong work allocation, the supervisor can move a compatible operation or share work across qualified stations. If the cause is a machine fault, maintenance owns the response and the plan must reflect the interrupted capacity. If the cause is missing material, the shortage remains visible rather than being converted into an operator performance loss. If a method change is required, industrial engineering must release it before new time comparisons are made.

WIP limits support this review, but they should not be treated as proof of flow by themselves. A line can remain below a WIP ceiling while one order waits untouched and another is repeatedly prioritised. The board therefore needs order or bundle age as well as quantity. It also needs the reason for any deliberate sequence exception, especially when an urgent order displaces work that was already promised.

The review should close with demonstrated capability, not with the act of rebalancing itself. Did the constraint move? Did good output increase without creating more rework? Did the oldest WIP fall? Did the new allocation remain within qualified skill coverage? Those questions connect the supervisor's intervention to delivery reliability. They also produce the evidence that the next monthly capacity review should use in place of nameplate assumptions.