Why PDCA Matters at the Individual Designer Level
PDCA — Plan, Do, Check, Act — is most commonly discussed as a quality management framework applied to production processes and organizational improvement programs. Its application to individual engineering work is less often discussed, but arguably more impactful. A designer who systematically improves their own practice — estimating more accurately, catching errors earlier, learning from every project — compounds their effectiveness over time in a way that no training course or tool upgrade can match.
This article is about applying PDCA at the level of a working mechanical designer’s day-to-day practice: how to plan design work with explicit assumptions, how to check your own work rigorously, and how to build improvement habits that make the next project better than the last.
Plan: Designing the Design Process, Not Just the Part
The plan phase in mechanical design is not just the schedule — it is the explicit identification of what decisions need to be made, what information is needed to make them, and what risks exist in the design approach. Engineers who skip this phase produce designs that have to be reworked because a critical constraint was discovered mid-design.
What a Good Plan Includes
- Design requirements with explicit success criteria: Not "the shaft must be strong enough" but "the shaft must achieve a safety factor of at least 2.0 under the maximum design load of 5 kN at the specified speed." Vague requirements cannot be verified.
- Identified unknowns and how to resolve them: Every design has open questions at the start. List them explicitly and identify the action required to resolve each one (vendor data request, analysis, prototype test). Unknown unknowns are unavoidable; unknown knowns — things you know you need to find out but haven’t written down — are not.
- Key decision points and their dependencies: What decisions must be made before detailed design can proceed? What information must be available for each decision? Identify the critical path through the design work, not just the total calendar time.
- Risk register: A brief list of the top three to five design risks, with a mitigation approach for each. This forces conscious engagement with what could go wrong and prevents the surprise that derails projects.
Do: Disciplined Execution with Self-Monitoring
The do phase is detailed design work — 3D modeling, analysis, drawing preparation. Continuous improvement in this phase comes from building checking habits into the work itself, not as a separate review step at the end.
In-Process Self-Checking Habits
- Check model mass and center of gravity at significant milestones: These gross metrics catch modeling errors (missing features, duplicate bodies) early, before the error propagates through the drawing set.
- Verify interface dimensions against mating assembly early: Don’t wait until the design is complete to check that mounting bolt patterns, shaft diameters, and connection points match. Check at the concept stage and again at the detailed design stage.
- Run the stress calculation before finalizing dimensions, not after: If the calculation shows the design is marginal, you want to know while the geometry can still change without revision cost. A stress check at the end of design that reveals a problem requires rework; the same check in the middle of design requires only a design adjustment.
- Read the drawing as a manufacturer: Before finalizing a drawing, ask: if I received this drawing with no other information, could I manufacture this part? Are all dimensions present? Are tolerances clear? Is the material specified? Could any view or section be misinterpreted?
Check: Formal and Informal Review
The check phase includes both formal design reviews and the less structured ongoing process of verifying that the design meets its stated requirements.
| Check Type | When | Purpose |
|---|---|---|
| Requirement traceability check | Before design review | Verify every requirement has a design response |
| Peer drawing review | Before release | Catch errors, ambiguities, missing information |
| Analysis review | After major design changes | Confirm safety factors still met after changes |
| Design-to-specification delta check | After design freeze | Identify any requirements that changed without design update |
| Post-build dimensional audit | After first article | Verify critical dimensions were achievable as specified |
The most underused check in engineering practice is the post-build dimensional audit on first articles. This check answers the question that every design process should close the loop on: did the design produce parts that actually met the dimensional specifications? Systematic deviations between design intent and as-built dimensions are the raw material for process improvement.
Act: Capturing Lessons and Updating Practice
The act phase is where PDCA generates compounding returns. Improvement actions taken after each project make the next project better. Engineers who skip this phase repeat the same avoidable errors indefinitely.
Practical Improvement Actions
- After every project, document two to three lessons learned — specific findings, not vague observations. "Specified keyway radius without checking achievability with standard broach tooling — resulted in stress concentration higher than analysis assumed" is useful. "Need to communicate better with manufacturing" is not.
- Update personal design checklists when a new error category is discovered. If you missed specifying a surface finish on a bearing bore and it caused a problem, add bearing bore surface finish to your pre-release checklist.
- Update hour estimates based on actual tracked hours. If a task consistently takes twice the estimated hours, the estimate is wrong. Correct it.
- Share lessons learned with the team. A team that learns individually but not collectively will rediscover the same failure modes in parallel.
FAQ
Q: How do you apply PDCA when projects move too fast for a structured approach?
Adapt the cycle to the pace rather than abandoning it. A ten-minute planning discussion at the start of a fast-moving task is still a plan phase. A five-minute self-check before sending a drawing is still a check phase. The value of the cycle is in the discipline of the thinking, not the time spent. Fast projects often benefit more from structured thinking, not less — the pace means there is less time to recover from errors that structured thinking would have prevented.
Q: What is the most common weakness in engineering PDCA cycles?
The act phase. Most design teams do some form of planning and checking. Very few have a consistent mechanism for capturing lessons and updating their process. Lessons learned sessions are held and then forgotten because there is no system for making the learning accessible at the point where it is needed — when the next similar design starts. Building the act phase into a searchable, accessible knowledge base is the most underinvested improvement action in most engineering organizations.
Q: How should I measure whether my design practice is actually improving?
Track metrics that reflect process quality, not just output. Useful metrics: revision rate per drawing (number of revisions issued after initial release), defect escape rate (design-related nonconformances per released drawing), estimation accuracy (actual hours versus estimated hours). These metrics reveal systemic patterns that subjective self-assessment misses. Review them quarterly and look for trends, not absolute values.
コメント