Utilisation decides whether installed capacity becomes good garments

The opened Engineering Reports article, published on 26 July 2025, reports Company X's OEE increasing from 42.91 percent to 52.27 percent with process monitoring devices and digital line balancing. It also reports Company Y moving from 59.47 percent to 67.30 percent. The arithmetic changes are 9.36 and 7.83 percentage points respectively. These are the article's case findings, not GFT results, and they do not establish the effect of either intervention in isolation.[S09]

The paired comparison in Exhibit 4 uses the same named measure on both rows. It shows the reported starting points and endpoints without applying them to another factory. The article demonstrates a relevant combination of monitoring and balancing. GFT's inference is that the combination deserves a locally designed test, not that a future line should inherit either endpoint as its financial-model assumption.[S09]

The ILO's article dated 1 June 2022 reports that Edget Garment improved line-balance efficiency from 24 percent to 75 percent and daily line production from 50 to 120 pieces following balancing work. The arithmetic changes are 51 percentage points and 70 pieces per day. This is a different measure and setting from the OEE cases. Neither the ratios nor the absolute levels should be combined into a common productivity benchmark.[S37]

The table in Exhibit 5 therefore keeps the efficiency and output rows distinct, including their units. The ILO describes examination of cycle time, takt time and operator requirements before a new balance was implemented. That sequence provides a practical reason to test measured work allocation. It does not establish what a greenfield line in Cameroon will deliver, how much it should cost or how long its improvement will take.[S37]

For GFT, the appraisal task is to convert a published mechanism into a testable local proposition. Specify the suspected loss, the operational change and the output expected to respond. If the proposal combines new data capture, supervisor practice and task reassignment, record that package explicitly. Do not subsequently attribute all improvement to the software alone. A useful comparison preserves the production context closely enough for the result to inform the next decision, while disclosing changes that prevent a clean interpretation.

GFT recommends a trial record that includes style and method, skill coverage, attendance, input readiness, quality status, available services and the definition of accepted output. Agree what would count as a material adverse result before beginning. This includes deterioration in working conditions, an increase in unresolved defects or a shift of unfinished work into a later period. The record should retain a trial that fails. Its value is the avoided repetition of a poor intervention, not a presentationally attractive success rate.

The result should support a bounded decision. An effective trial can justify retaining a method under the tested conditions and planning a further test where conditions differ. It should not automatically justify expanding every line, replacing the factory's system or rewriting the entire production forecast. GFT recommends attaching the decision scope directly to the evidence: what the test established, what it did not test and what new condition would require a fresh check. This preserves useful case evidence without turning it into a promise.

GFT proposes carrying the performance-management remit of its Cameroon textile programme into a practical operating record: output, defects and downtime captured from the first production shift, with hourly review where it supports intervention. This is a recommendation for implementation, not a claim that a factory has already adopted the system or achieved a result. The record should connect an observed deviation to a named operating role, a chosen action and a later check at the accepted-output boundary.

Start with definitions that a supervisor and operator can use together. State what counts as completion, when a hold is released, which minutes belong to production and how overlapping stops are handled. Record unresolved causes as unresolved rather than force-fitting them into a convenient category. GFT recommends checking a sample of recorded events against the physical line before interpreting a trend. A missing event is a measurement problem; it should not silently become evidence that the corresponding loss did not occur.

For the review itself, GFT recommends beginning with the gap between authorised output and accepted output. Ask which restriction explains it and whether the available observations support that explanation. Choose a corrective action proportionate to the evidence. Record who will act and how the next review will establish whether the restriction changed. Avoid declaring closure because a maintenance ticket was signed or a dashboard turned green. Closure belongs to the agreed operating consequence, which may require more observation.

An illustrative decision record could concern recurring material waiting ahead of the constraint. The proposed action might be a defined pre-shift readiness check and a controlled buffer of qualified components. The acceptance condition would concern sustained accepted output and absence of uncontrolled WIP growth, rather than faster material issue alone. If the same record instead shows a quality hold, the response should follow the defect mechanism. The data system earns its place by supporting the distinction and preserving the outcome of the action taken.

Automation should follow a written event contract. GFT proposes stating which physical event a sensor or operator entry represents, when it receives a timestamp and how an error can be corrected. A machine pulse should not become a good-garment count without a demonstrated connection to the relevant operation and acceptance event. Preserve manual reconciliation during the trial. If a network interruption occurs, distinguish missing telemetry from stopped production. The detailed governance and reference architecture sit elsewhere in the paper library; the operating requirement is interpretable evidence.

The paper record should remain usable when a shift changes. GFT proposes a short handover that separates unresolved production restrictions from unresolved measurement questions. A production restriction needs an operational owner; a measurement question needs a check before it becomes part of a performance conclusion. Carry forward the evidence and the next action, rather than only the latest percentage. When the team corrects an earlier count, preserve the correction and its reason. That makes the next shift's review a continuation of the same investigation and allows the sponsor to reconstruct how a claimed operating improvement was established.

Before accepting a claimed improvement, the sponsor should request a concise reconciliation: planned output, accepted output, key losses, unfinished stock movement and any deterioration in quality or working conditions. Compare like production boundaries and disclose changed assumptions. This chapter's illustrative arithmetic is a method demonstration and is not investment, legal or tax advice. A project decision needs current, project-specific operating and commercial evidence. The remaining chapters develop the loss tree, work measurement, balancing, flow and investment controls needed to supply it.