11.2 Plan-Do-Check-Act (PDCA) & Solution Selection

Plan-Do-Check-Act (PDCA) & Solution Selection

Quick Answer: The Plan-Do-Check-Act (PDCA) cycle, also known as the Deming wheel or Shewhart cycle, is an iterative scientific four-stage management method for continuous process improvement. In the Improve phase, teams brainstorm countermeasures addressing root causes verified in Analyze, evaluate them using weighted Solution Selection Matrices, and test selected solutions via controlled small-scale pilots during the "Do" stage. Results are evaluated against baseline data in "Check/Study", and verified improvements are standardized and scaled in "Act". Independent CSSYB study guide by OpenExamPrep.

---

Foundations of the PDCA / PDSA Cycle

The Plan-Do-Check-Act (PDCA) cycle provides the scientific foundation for structured quality improvement. The framework traces to Dr. Walter A. Shewhart, the Bell Telephone Laboratories statistician who pioneered statistical process control and, in 1939, described quality control as a cycle of specification, production, and inspection. Dr. W. Edwards Deming taught a four-step version of that cycle to Japanese managers in 1950, and Japanese practitioners turned it into the Plan-Do-Check-Act cycle.

Deming frequently termed the model PDSA (Plan-Do-Study-Act), deliberately substituting "Study" for "Check." Deming argued that "check" implied a passive inspection, whereas "study" emphasized active statistical analysis, reflection on empirical findings, and understanding the root mechanisms of process behavior. In Six Sigma, PDCA serves as the operational engine of the Improve phase, moving teams from diagnostic hypotheses to verified, scalable solutions.

Rather than a closed circle, PDCA operates as an upward spiral of continuous learning. Each cycle deepens process understanding, eliminates operational variation, and elevates the baseline standard for future performance.

---

The Four Stages of PDCA in the Improve Phase

Applying PDCA during solution implementation requires disciplined execution across four sequential stages:

1. Plan: The team reviews the vital few input variables ($X$) confirmed during the Analyze phase and establishes clear, quantifiable SMART improvement targets. Members generate candidate countermeasures and design a detailed pilot implementation plan, including sample sizes, trial durations, and control metrics.
2. Do: The team executes the selected countermeasures on a small, controlled scale (such as a single workstation, one operating shift, or an isolated pilot cell). Running a limited pilot allows the team to observe real-world performance and collect empirical data without introducing enterprise-wide operational risk.
3. Check (Study): The team collects pilot performance data and evaluates it against pre-improvement baselines. Members assess capability gains ($C_p, C_{pk}$), defect reductions (DPMO), and cycle time changes. The team also checks for unintended consequences, such as bottleneck migrations or operator ergonomics issues.
4. Act: If the pilot achieves performance goals, the team standardizes the countermeasure enterprise-wide by updating Standard Operating Procedures (SOPs), installing mistake-proofing (Poka-Yoke), and training staff across all shifts. If the pilot falls short, the team analyzes the gaps, adjusts hypotheses, and initiates the next iterative PDCA cycle.

---

Solution Generation & Brainstorming Countermeasures

During the Analyze phase, teams isolate the root causes that drive quality variation. In the Improve phase, focus shifts to generating innovative countermeasures that eliminate or control those root causes.

To avoid premature convergence on obvious or biased fixes, teams utilize structured ideation techniques:

• Brainwriting (6-3-5 Method): Six participants write three ideas on worksheets within five minutes, passing them silently across six rounds. This produces 108 distinct ideas while eliminating loud-voice domination and executive bias.
• Nominal Group Technique (NGT): Participants silently generate ideas, post them to a visual board, discuss each concept for clarity, and vote silently using weighted scoring to prioritize top candidates.
• Benchmarking & Analogy Thinking: Teams examine how analogous industries solve similar handling or assembly challenges, adapting proven concepts to their operational environment.

---

The Solution Selection Matrix (Prioritization Matrix)

When multiple candidate countermeasures emerge, unstructured debates often lead to political compromise or executive preference. A Solution Selection Matrix (or Prioritization Matrix) establishes mathematical objectivity by evaluating alternative solutions against weighted operational criteria.

Evaluation Criteria and Weightings

A balanced matrix evaluates both customer-facing quality gains and organizational constraints:

• Impact on Critical-to-Quality (CTQ) / Defect Reduction (30% - 40%): Degree to which the solution eliminates the verified root cause and reduces process defects.
• Implementation Cost / Capital Expense (20% - 30%): Total capital expense, tooling costs, software licenses, and physical installation labor.
• Technical & Operational Feasibility (15% - 25%): Ease of operating and maintaining the solution using existing staff and equipment.
• Time to Implement / Speed to Value (15% - 25%): Duration required to design, procure, test, and fully deploy the solution.

Each solution receives a score (typically 1 to 5) per criterion. The weighted composite score is calculated using the formula:

$$\text{Composite Score} = \sum (\text{Rating}_i \times \text{Weight}_i)$$

Worked Example: Packaging Line Countermeasures

Consider three candidate solutions designed to eliminate missing product manuals:

• Solution A: Two-person manual inspection checklist.
• Solution B: Inline optical barcode scanner with automated diverter gate.
• Solution C: Fully automated robotic packaging cell.

Analysis: Solution A is inexpensive but fails to reliably eliminate defects. Solution C provides perfect quality but carries excessive capital costs and lengthy lead times. Solution B achieves the highest composite score (4.35), balancing superior defect prevention with affordable cost and rapid execution.

---

Pilot Testing: Controlled Deployment & Risk Mitigation

Before committing capital and organizational resources to full-scale deployment, teams execute a controlled pilot test during the "Do" stage. A pilot test is a small-scale implementation conducted in an operational sandbox—such as a single production line, one shift, or a designated customer segment.

Strategic Objectives of a Pilot Test

1. Empirical Verification: Proves that the solution reduces defects under real operating conditions, including shift changes, raw material lots, and operator variation.
2. Uncovering Side Effects: Identifies unforeseen bottlenecks, ergonomic friction, or software latency before they disrupt plant-wide operations.
3. Refining Standard Work: Exposes ambiguities in draft work instructions, allowing teams to clarify procedures and refine visual controls based on operator feedback.
4. Building Frontline Champions: Pilot operators become credible peer advocates, easing cultural resistance when the change rolls out across the broader organization.