Reduce Bottling Line Changeover Time by Finding the Limiting Tasks

Choose a transition that matters to the production plan, with the same outgoing and incoming revisions, crew basis and line-end measurement point. Preserve timestamps for each task and for first accepted and sustained accepted new output. Record rejects, material use and unusual interventions beside the time, so a faster restart does not hide extra waste.

• Do not mix time from unrelated SKU pairs into one improvement claim.
• Record both elapsed minutes and staffing; a shorter window with extra people is a different operating basis.
• Use existing approved procedures. Observation or timing must not interfere with safe operation.

Check whether parts identification, document checks and material preparation can be completed before the transition without entering restricted areas or disturbing production. Work that requires isolation, clearance, approved hygiene steps or first-off checks stays under the required procedure. Moving work outside the measured stop does not eliminate its labor or material cost.

• Assign a part/recipe revision and an owner to each preparation task.
• Parallel work is possible only when access, staffing, machine dependencies and safety procedures allow it.
• Record waiting for tools, materials, approval or a preceding task instead of hiding it inside “adjustment”.

Independent illustrative sequence: minute 0 is the last accepted outgoing unit at the agreed line-end point. Approved clearance takes minutes 0–6. Mechanical conversion then takes 6–22, while a separate permitted recipe task runs 6–14. First-off checks run 22–28 and verified recovery runs 28–38, ending at the agreed sustained-output condition. The tasks total 6 + 16 + 8 + 6 + 10 = 46 minutes, but the elapsed window is 6 + max(16,8) + 6 + 10 = 38 minutes. This is a dependency example, not an operating instruction or measured customer result.

• The transition window is not necessarily 38 minutes of zero output: accepted new units may appear during recovery. Keep their counts and the actual stop/running states for the production record.
• Reducing the recipe task from 8 to 4 minutes alone leaves the 38-minute finish unchanged.
• If an approved change reduces the mechanical task from 16 to 11 minutes with all other conditions unchanged, the hypothetical window becomes 33 minutes.
• The five minutes of modeled opportunity are not guaranteed extra saleable production; confirm repeatability, quality and downstream recovery.

Repeat the same transition after one documented change and show the observed range, not only the fastest trial. Keep changes in crew, materials, quality checks and utilities visible. Confirm whether accepted line-end output returns sooner and whether the loss moved to a different task or machine.

• If preparation or waiting governs the finish, evaluate that evidence before buying automation for an unrelated task.
• If a verified mechanical limitation remains, compare an engineered change or replacement against the affected transition and required output.
• Do not bypass guards, interlocks, isolation requirements or hygiene/quality checks to obtain a shorter time. Send a non-confidential task timeline and retained-equipment details for review.