ORDREL / LEARN · ENGLISH
A3 Problem Solving: A Worked Manufacturing Example
An A3 is a concise problem-solving narrative: the current condition, the gap, the evidence-based explanation, the proposed countermeasure and the plan to check it. The page format supports thinking; filling every box is not proof of improvement.
Choose a bounded problem and a useful measure
Name the process, customer requirement, time window and owner. Separate background from the observed gap. Define the start and end of the measure so that the baseline and follow-up are comparable. Record containment separately from the work needed to prevent recurrence.
A target is a proposed decision threshold, not an achieved result. State why it matters to the operation. If the constraint, cost or customer consequence has not been established, say so instead of inventing savings.
Build a fictional changeover A3
Fictional manufacturing teaching example. These figures are not ORDREL customer results.
| Element | Teaching example |
|---|---|
| Current condition | Four comparable changeovers take 42, 38, 45 and 35 minutes from last good unit to first approved good unit. |
| Gap and target | Mean is 40 minutes; the team proposes testing a 30-minute mean under the same boundary. |
| Evidence | Direct observation separates search, preparation, adjustment and approval; causes remain hypotheses until checked. |
| Candidate countermeasure | Stage verified tooling and materials before the planned stop, using the existing approval process. |
| Trial | Use a limited, approved trial; record the same time boundary, quality outcome and any transferred workload. |
| Follow-up | Review results, unintended effects and repeatability before updating the standard method. |
Baseline mean = (42 + 38 + 45 + 35) ÷ 4 = 40 minutes. A fictional trial records 31, 29, 32 and 28 minutes: mean 30 minutes. The observed difference is ten minutes, or 25% of the baseline mean. These small samples illustrate the calculation; they do not establish statistical stability or causation.
Make the causal argument visible
Suppose the team observes repeated searches for a fitting after the machine stops. Record the actual delay and why the needed item was not available. Do not assume that every minute above the target is search time. Check whether advance staging is feasible, safe and truly removes delay rather than moving it to an unmeasured person or shift.
A countermeasure should address the supported mechanism. A generic instruction to “work faster” does not explain what changes. Include the expected observation if the countermeasure works and the condition under which it should be stopped.
Check quality and transferred work, not only time
Faster changeovers are not useful if first-run defects, ergonomic load or approval failures increase. Track those guardrails alongside elapsed time. Record mix, equipment condition and any concurrent change that could explain a difference.
If a result fails to repeat, revise the explanation or trial design. Do not edit the target to make the worksheet look successful. A follow-up date, accountable owner and explicit unresolved questions are more useful than an unsupported “closed” status.
Choose the working record or the separate playbook
The A3 workspace provides a browser-based record for the problem-solving narrative. Start with a narrow case and keep evidence references free of personal or customer information. Save/export availability follows the current plan and account state.
The A3 Word playbook is a separate one-time purchase. It is not a promise of consulting, certification or automatic root-cause validation. Use Fishbone or 5 Whys when the investigation needs detail that would make the A3 unreadable.
- Does an A3 need to fit one physical sheet? The discipline is concise connected reasoning; do not omit essential evidence merely to fit a template.
- Is a before/after average enough? It is a descriptive start, not proof. Examine comparability, variability and other changes.
- Who is the author here? This is ORDREL teaching material, not a customer case study.
Sources and method notes
Sources support the method definitions. Examples, interpretation and tool instructions are ORDREL teaching material, not an endorsement by the referenced organizations. See methods and corrections.