Introduction: Your Monofocal SPC Has Two Charts. Your EDOF SPC Needs Eight. Here’s How to Organize Them.
The process engineer responsible for monofocal IOL SPC runs two control charts: X-bar/R on sphere power and X-bar/R on cylinder. The process is stable. The charts are boring. This is exactly what good SPC looks like-controlled variation within established limits, no patterns, no action required.
Then EDOF production launches. The measurement system generates through-focus plateau width, minimum MTF within the designed range, plateau center position, plateau symmetry, multi-aperture performance, SA coefficients (Z₄⁰, Z₆⁰, Z₈⁰), the Z₄⁰/Z₆⁰ ratio, total parasitic RMS, and power and cylinder from the standard protocol. That is twelve parameters from a single lens measurement. Charting all twelve produces a wall of control charts that no operator can monitor in real time and no process engineer can interpret efficiently.
EDOF SPC is not about charting every available parameter. It is about selecting the right parameters at the right level of the monitoring hierarchy, setting control limits that reflect the actual manufacturing process, and interpreting out-of-control signals through the lens of EDOF-specific manufacturing physics.
This article provides the practical setup: the three-tier parameter hierarchy that organizes twelve parameters into a manageable monitoring system, the method for calculating control limits when you have no historical EDOF data, and the out-of-control pattern dictionary that translates statistical signals into manufacturing corrective actions.
The Three-Tier Parameter Hierarchy: Primary, Diagnostic, and Monitoring
The twelve parameters available from EDOF measurement are not equally important for SPC. Some are outcome metrics that directly represent the lens’s clinical performance. Others are diagnostic metrics that explain why an outcome metric changed. Still others are background metrics that confirm nothing unexpected is happening but do not drive daily process decisions.
Organizing these parameters into a hierarchy determines which parameters get their own real-time control chart (primary tier), which parameters are plotted and reviewed when a primary chart signals an issue (diagnostic tier), and which parameters are archived for trend analysis but not actively monitored in real time (monitoring tier).
Tier 1: Primary charts – the outcome metrics
These are the parameters that directly represent the EDOF performance the patient experiences. They are the functional output of the manufacturing process. If these are in control, the process is producing lenses that work. If these go out of control, lenses that don’t work may be shipping.
Chart 1: Through-focus plateau width. This is the single most important EDOF quality metric. It directly represents the extended range of vision that defines the product. An X-bar chart on plateau width is the EDOF equivalent of the power chart for monofocal. When this chart signals a shift, the EDOF performance characteristic is compromised.
Chart 2: Minimum MTF within the designed range. A wide plateau with a deep dip somewhere within the range is clinically worse than a slightly narrow plateau without dips. This parameter catches the contrast dead zones that plateau width alone may miss.
Chart 3: Power at nominal focus. Standard monofocal parameter, retained for EDOF. The lens must still focus at the correct distance.
Three primary charts. These are displayed prominently, updated in real time, and monitored by the QC operator on every shift.
Tier 2: Diagnostic charts – the root cause indicators
When a primary chart signals an out-of-control condition, the diagnostic charts explain why. These parameters are plotted continuously but reviewed in detail only when a primary signal triggers investigation.
Chart 4: Z₄⁰ (primary SA). Deviation in Z₄⁰ shifts the plateau center and changes the plateau width. This is the first diagnostic parameter to check when plateau width drifts. Mapping from Z₄⁰ deviation to manufacturing cause is straightforward: conic constant error in the aspheric profile.
Chart 5: Z₆⁰ (secondary SA). Z₆⁰ is the efficiency multiplier in the EDOF SA recipe. Under-delivery of Z₆⁰ is the most common cause of plateau narrowing that is not explained by Z₄⁰ deviation. The Zernike coefficient interpretation framework maps Z₆⁰ deviation to 6th-order aspheric coefficient error in the CNC tool path.
Chart 6: Z₄⁰/Z₆⁰ ratio. Individual SA coefficients can each be within limits while their ratio drifts out of the optimal range. The ratio chart catches interaction effects that individual coefficient charts miss. This is unique to EDOF SPC-monofocal has no equivalent.
Chart 7: Total parasitic RMS. The root-sum-square of all non-designed Zernike terms (coma, trefoil, astigmatism). Elevated parasitic RMS degrades the EDOF plateau through interaction with designed SA. A single composite chart is more efficient than separate charts for each parasitic term.
Four diagnostic charts. Reviewed when primary charts signal an issue. Collectively, they identify whether the root cause is in the designed SA profile (Charts 4–6) or in manufacturing-introduced aberrations (Chart 7).
Tier 3: Monitoring parameters – background tracking
These parameters are recorded and available for trend analysis but do not require dedicated real-time control charts.
Plateau center position. Important for acceptance criteria but changes slowly-typically correlated with Z₄⁰, which is already charted at Tier 2.
Plateau symmetry. Correlated with coma (Z₃¹), which is captured in the total parasitic RMS at Tier 2.
Multi-aperture performance. Verified on sampled lenses via the IOLA MFD; meaningful for batch disposition but changes too slowly to warrant a per-lens real-time chart.
Z₈⁰ (tertiary SA). Typically too noisy (wide manufacturing distribution) to produce actionable SPC signals. Monitor for trends; do not react to individual data points.
Cylinder. Standard parameter from monofocal SPC. Retained for completeness.
Table 1: EDOF SPC Parameter Hierarchy
| Parameter | Tier | Chart Type | Update Frequency | What It Catches | Action When Out of Control |
| Plateau width | 1 – Primary | X-bar / R (or I-MR) | Every lens or every batch | EDOF range degradation; the single metric most correlated with patient outcome | Hold batch; investigate using Tier 2 diagnostics |
| Minimum MTF within range | 1 – Primary | X-bar / R (or I-MR) | Every lens or every batch | Contrast dead zones within the plateau; catches dips that width metric misses | Hold batch; check if dip is in diffractive ripple region (design) or random position (defect) |
| Power at nominal focus | 1 – Primary | X-bar / R | Every lens | Standard power accuracy; retained from monofocal SPC | Standard power correction protocol |
| Z₄⁰ (primary SA) | 2 – Diagnostic | I-MR | Every lens or batch sample | Aspheric profile accuracy; primary contributor to plateau width | Adjust conic constant; check tool path |
| Z₆⁰ (secondary SA) | 2 – Diagnostic | I-MR | Every lens or batch sample | Higher-order profile accuracy; efficiency multiplier deviation | Adjust 6th-order aspheric coefficient; check central zone machining |
| Z₄⁰/Z₆⁰ ratio | 2 – Diagnostic | I-MR | Every lens or batch sample | SA balance interaction; catches ratio drift when individual terms are within limits | Identify which SA term is drifting faster; correct the dominant contributor |
| Total parasitic RMS | 2 – Diagnostic | I-MR | Every lens or batch sample | Manufacturing alignment and fixturing quality; composite manufacturing noise metric | Decompose: if coma dominant → alignment. If trefoil → clamping. If astigmatism → stress. |
| Plateau center position | 3 – Monitoring | Trend chart (weekly) | Batch average | Plateau position drift on defocus axis; correlated with Z₄⁰ shift | Review if Z₄⁰ chart shows trend; otherwise archive for quarterly review |
| Plateau symmetry | 3 – Monitoring | Trend chart (weekly) | Batch average | Asymmetric degradation; correlated with coma in parasitic RMS | Review if parasitic RMS chart shows elevation; otherwise archive |
| Z₈⁰ (tertiary SA) | 3 – Monitoring | Trend chart (weekly) | Batch average | Fine plateau edge shaping; high manufacturing noise makes per-lens SPC impractical | Monitor for sustained trends only; individual outliers are expected |
| Cylinder | 3 – Monitoring | X-bar / R (existing) | Every lens | Standard cylinder accuracy; retained from monofocal SPC | Standard cylinder correction protocol |
Calculating Control Limits When You Have No Historical EDOF Data
Standard SPC control limits are calculated from 25–30 subgroups of stable process data. For a new EDOF launch, this data does not exist yet. The process engineer faces a bootstrapping problem: control limits require stable production data, but production has not yet started.
Phase 1: Provisional limits from pre-production data (weeks 1–4)
During the acceptance criteria development period, the process engineer measures 30–50 pre-production lenses across multiple batches. This dataset provides the initial estimate of the manufacturing distribution for each parameter.
Provisional control limits are calculated from this dataset using standard X-bar/R or I-MR formulas. The center line is the grand mean of the pre-production data. The control limits are set at ±3σ (calculated from the within-subgroup variation for X-bar/R charts, or from the moving range for I-MR charts).
These provisional limits are explicitly labeled as such. They are based on a limited dataset and may not accurately represent the full range of stable production variation. They serve one purpose: establishing a baseline so that SPC monitoring can begin from the first day of production, rather than running blind for the first month.
Phase 2: Revised limits from early production data (weeks 5–12)
After 4–8 weeks of production, the dataset grows from 30–50 pre-production lenses to 200–500 production lenses. The control limits are recalculated from the production data, excluding any periods where the process was demonstrably out of control (documented corrective actions, known upsets, startup transients).
The revised limits are typically tighter than the provisional limits because the production data better represents normal operating variation (pre-production batches may include startup variability that is not representative of steady-state production). If the revised limits are wider than the provisional limits, the process may be less stable than pre-production suggested-an important early warning.
Phase 3: Validated limits from mature production data (month 3+)
After 3 months of stable production, the control limits are validated against a dataset of 500 or more lenses. The process capability indices (Cpk) are calculated for each primary and diagnostic parameter. The control limits become the official SPC specification, documented in the quality system and subject to formal change control.
This three-phase approach-provisional, revised, validated-ensures that SPC monitoring begins on day one while progressively improving the accuracy of the control limits as production data accumulates.
Table 2: Control Limit Calculation by Phase
| Phase | Data Source | Limit Calculation | Typical Precision | Action |
| 1: Provisional (weeks 1–4) | 30–50 pre-production lenses from ≥3 batches | I-MR: CL = X̄, UCL/LCL = X̄ ± 2.66 × MR̄. X-bar/R (subgroup n=5): CL = X̄̄, UCL/LCL = X̄̄ ± A₂ × R̄ | 95% CI for σ estimate: ±25–35%. Limits are approximate. | Label as “Provisional.” Start monitoring. Do not use for formal process capability claims. |
| 2: Revised (weeks 5–12) | 200–500 production lenses (excluding known upsets) | Same formulas; recalculated from production data. Exclude out-of-control periods from limit calculation. | 95% CI for σ estimate: ±10–15%. Limits are reliable. | Replace provisional limits. Calculate preliminary Cpk. Identify any parameter with Cpk < 1.0. |
| 3: Validated (month 3+) | 500+ production lenses from stable, verified-in-control production | Final calculation. Verify normality assumption. If non-normal, consider transformed limits or non-parametric approach. | 95% CI for σ estimate: ±7–10%. Limits are definitive. | Document as official SPC specification. Subject to formal change control. Calculate validated Cpk. |
[Note: Control limit formulas follow standard SPC methodology (Montgomery, Introduction to Statistical Quality Control). For I-MR charts, the 2.66 factor assumes normally distributed data. For EDOF parameters with skewed distributions (e.g., total parasitic RMS, which is Rayleigh-distributed), transformed limits or distribution-specific approaches may be needed. Consult your statistician if the normality assumption is substantially violated.]
Data Flow: From Measurement to Chart
The SPC data pipeline must connect the measurement system to the control charts with minimal manual intervention. Manual data entry introduces transcription errors and delays that undermine real-time monitoring.
100% inspection flow (Tier 1 parameters)
For facilities using the IOLA MP for 100% batch inspection, every lens measurement generates the power and cylinder data that feeds Tier 1 Chart 3 (power) and Tier 3 (cylinder). The IOLA MP’s automatic data export-in TXT, Excel, or SQL formats-sends measurement results directly to the SPC system. Each lens appears on the control chart within seconds of measurement.
If through-focus parameters (plateau width, minimum MTF) are computed on every lens-possible when the measurement system captures wavefront data-Tier 1 Charts 1 and 2 are also populated from the 100% inspection flow.
Sampled through-focus flow (Tier 1 and 2 parameters)
For facilities using the IOLA MFD for through-focus verification on sampled lenses per batch, the through-focus parameters and SA coefficients from each sampled lens feed the Tier 1 through-focus charts and the Tier 2 diagnostic charts. The MFD’s single 9-second measurement generates all Tier 1 and Tier 2 parameters simultaneously.
The sampling rate determines the chart type. If 5 lenses per batch are measured, X-bar/R charts with subgroup size n=5 are appropriate. If one lens per batch is measured, Individual-Moving Range (I-MR) charts are used.
Automated rule checking
Modern SPC software can apply Western Electric rules (or Nelson rules) automatically and alert the operator when a pattern is detected. For EDOF SPC, configure the following rule set as a minimum:
Rule 1: One point beyond 3σ (standard out-of-control). Rule 2: Two of three consecutive points beyond 2σ on the same side (shift warning). Rule 5: Six consecutive points trending in the same direction (drift detection). Rule 7: Fifteen consecutive points within 1σ on either side (reduced variation-may indicate a gauge issue or data recording problem).
When a rule triggers on a Tier 1 chart, the system automatically highlights the corresponding Tier 2 charts for that time period. The process engineer opens the diagnostic view and reads the SA coefficients and parasitic RMS to determine whether the issue is in the designed profile or in manufacturing-introduced aberrations.
What EDOF-Specific Out-of-Control Patterns Mean
SPC patterns on monofocal charts have straightforward interpretations: a power shift means the curvature changed. EDOF SPC patterns are more nuanced because the through-focus performance depends on multiple interacting parameters.
Table 3: EDOF SPC Out-of-Control Pattern Dictionary
| Pattern on Tier 1 Chart | Tier 2 Diagnostic Finding | Manufacturing Root Cause | Corrective Action |
| Plateau width: sudden downward shift (Rule 1 or 2) | Z₆⁰ shifted below target; Z₄⁰ unchanged | 6th-order aspheric coefficient error; likely CNC tool path update error, tool change, or thermal transient affecting central zone machining | Verify current tool path version. Check tool condition. Measure central zone profilometry. Adjust 6th-order coefficient. |
| Plateau width: gradual downward trend (Rule 5) | Z₄⁰ trending downward; Z₆⁰ stable | Progressive tool wear changing the conic constant; thermal drift over the production shift; slow material RI change across batch | Check tool wear indicator. Monitor environmental temperature log. Verify material lot documentation. |
| Plateau width: sudden shift upward (wider than expected) | Z₄⁰ and Z₆⁰ both increased | Over-correction: too much SA introduced. Possible: wrong tool path loaded for a higher-power lens, or material RI higher than nominal | Verify product code vs loaded tool path. Check material lot RI. Wider plateau is not automatically good-peak MTF may be degraded. |
| Minimum MTF: sudden drop | Parasitic RMS elevated; Z₃¹ (coma) dominant | Surface decentration: front/back surface misalignment increased, creating coma that interacts with designed SA to deepen through-focus dip | Realign collet or blocking tool. Measure decentration directly. Re-verify after alignment correction. |
| Minimum MTF: gradual degradation | Parasitic RMS stable; SA coefficients stable; wavefront residual shows increasing periodic structure | Mid-spatial frequency errors increasing; progressive tool wear creating deeper tool marks that scatter light and reduce MTF within the plateau | Inspect tool under microscope. Replace tool. Reduce feed rate. Verify improvement on next batch. |
| Z₄⁰/Z₆⁰ ratio: drift while individual terms are in-control | Z₄⁰ drifting slightly up while Z₆⁰ drifting slightly down (or vice versa); individual charts show no rule violation | Coupled form error: the same process change (e.g., temperature) affects Z₄⁰ and Z₆⁰ in opposite directions. Each change is small but the ratio shift is significant. | This is the pattern that only the ratio chart catches. Correct the process condition causing the coupled drift (typically thermal or tool geometry). |
| Parasitic RMS: step increase | Decompose: if Z₃³ (trefoil) dominant → clamping. If Z₂² (astigmatism) dominant → stress. If Z₃¹ (coma) → alignment. | Fixturing change, chuck jaw wear, blocking adhesive batch change, or environmental vibration increase | Match dominant parasitic term to mechanical cause per the decomposition. Correct specific fixturing or alignment parameter. |
| Power: standard shift (same as monofocal) | Z₄⁰ may also shift (aspheric profile and power are coupled) | Curvature change affecting both power and SA simultaneously | Standard power correction. Additionally verify Z₄⁰ returned to target after correction-power correction in EDOF must not inadvertently change the SA profile. |
| All Tier 1 charts show reduced variation (Rule 7) | All Tier 2 charts also show narrower ranges | Possible measurement system issue: calibration drift, software update, or gauge malfunction creating artificially consistent readings | Run gauge verification with reference lens. If gauge is verified, the process genuinely improved (document as such). If gauge fails, recalibrate. |
The Ratio Chart: The EDOF-Specific Innovation
The Z₄⁰/Z₆⁰ ratio chart (Tier 2, Chart 6) deserves special attention because it has no equivalent in monofocal SPC and catches failure modes that no other chart detects.
The through-focus plateau shape is determined by the interaction between SA orders, not by their individual values. Two lenses with identical Z₄⁰ and different Z₆⁰ produce different plateaus. The ratio captures this interaction in a single number.
The ratio chart catches a specific failure mode: coupled drift. When a process condition (temperature, tool geometry, material RI) changes, it often affects Z₄⁰ and Z₆⁰ simultaneously but in different directions or proportions. Each individual coefficient may remain within its control limits because the drift in each is small. But the ratio between them shifts enough to change the plateau shape.
Example: Z₄⁰ drifts from -0.150µm to -0.145µm (within limits). Z₆⁰ drifts from +0.080µm to +0.072µm (within limits). The ratio shifts from -1.875 to -2.014-a 7% change. The individual charts show nothing. The ratio chart signals a drift. The through-focus plateau has already changed shape.
Setting up the ratio chart: compute the ratio for each measured lens. Plot on an I-MR chart. Calculate control limits from the ratio data using the same three-phase approach. The ratio distribution is often approximately normal (because it is the quotient of two approximately normal variables with the same sign), but verify this with the actual data.
The ratio chart is the single most valuable addition that EDOF SPC makes to the monofocal SPC framework. If you implement only one new chart for EDOF, implement this one.
Practical Implementation: Week-by-Week Setup Guide
Week 1: Identify data sources and configure export.
Map which measurement system generates which parameters. Configure automatic data export from the IOLA MP (power, cylinder on every lens) and the IOLA MFD (through-focus, SA coefficients, parasitic RMS on sampled lenses). Verify data format compatibility with the SPC software.
Week 2: Calculate provisional limits from pre-production data.
Using the 30–50 pre-production lenses measured during acceptance criteria development, compute the mean and range/moving range for each parameter. Calculate provisional control limits for all 7 active charts (3 primary + 4 diagnostic). Configure the SPC software with these provisional limits.
Week 3: Begin monitoring with provisional limits.
Production starts. Every lens or batch sample populates the control charts. Operators monitor Tier 1 charts in real time. Process engineer reviews Tier 2 charts at end of each shift. Flag any out-of-control signals for investigation but use engineering judgment-provisional limits may be too tight or too loose.
Weeks 4–8: Accumulate production data and refine.
Collect 200–500 lens data points. Review each parameter’s distribution: is the mean centered on target? Is the distribution normal? Are there any parameters where the provisional limits are clearly inappropriate (e.g., all data points are near one limit)?
Week 8–12: Calculate revised limits from production data.
Recalculate control limits from the stable production data. Replace provisional limits with revised limits in the SPC system. Calculate preliminary Cpk for each parameter. Identify any parameter with Cpk < 1.33-this is the constraint parameter that limits process capability.
Month 3+: Validate limits and formalize.
With 500+ data points, validate the control limits. Verify the normality assumption for each parameter (particularly the ratio chart and total parasitic RMS, which may be non-normal). Document the validated SPC specification as part of the quality system. Establish the ongoing review cadence: quarterly recalculation of Cpk, annual review of control limit adequacy.
Common Pitfalls in EDOF SPC Setup
Pitfall 1: Charting everything
Twelve parameters on twelve charts creates visual overload. Nobody monitors twelve charts effectively. The three-tier hierarchy exists to prevent this. Three primary charts for real-time monitoring. Four diagnostic charts for triggered investigation. Five monitoring parameters for background tracking. Seven active charts total-manageable for one operator.
Pitfall 2: Using specification limits as control limits
Specification limits (from acceptance criteria) define what product is acceptable to ship. Control limits define what the process normally does. These are different numbers serving different purposes. If the process mean is centered and the process is capable, the control limits are tighter than the specification limits. If the control limits are wider than the specification limits, the process is not capable of meeting the specification-a capability problem, not an SPC problem.
Pitfall 3: Ignoring the ratio chart
Individual SA coefficient charts are necessary but not sufficient. The ratio chart catches the coupled drift failure mode that individual charts miss. Omitting the ratio chart is equivalent to running monofocal SPC on an EDOF line-it monitors the parts but not the interaction.
Pitfall 4: Applying monofocal rule interpretation to EDOF signals
In monofocal SPC, a power shift means a curvature change. The diagnosis is one-to-one. In EDOF SPC, a plateau width shift can result from Z₄⁰ change, Z₆⁰ change, Z₄⁰/Z₆⁰ ratio change, elevated parasitic RMS, or mid-spatial frequency errors. The diagnostic pathway is one-to-many. Jumping to a corrective action based on the Tier 1 signal alone-without checking the Tier 2 diagnostics-risks correcting the wrong parameter.
Pitfall 5: Setting limits too early with too little data
Provisional limits from 30 lenses have wide confidence intervals on the σ estimate (±25–35%). These limits will produce false alarms (type I errors) or missed signals (type II errors) at higher rates than validated limits. Accept this during Phase 1 and commit to recalculating at Phase 2 and Phase 3. Do not treat provisional limits as permanent.
From SPC to Process Improvement: Closing the Loop
SPC is a monitoring system. It detects change. It does not fix change. The value of EDOF SPC lies in how quickly and accurately the detected signal is translated into a manufacturing correction.
The three-tier hierarchy accelerates this translation. When Tier 1 signals an issue, Tier 2 provides the diagnosis within seconds-because the diagnostic charts are already plotted. The process engineer does not need to remeasure, re-analyze, or re-compute. The power map and Zernike data from the same measurement that triggered the SPC signal contain all the diagnostic information needed.
The corrective action is executed. The next batch is measured. The control charts confirm whether the correction was effective. If the primary charts return to control, the issue is resolved. If not, the diagnostic charts reveal whether the correction addressed the right parameter.
This feedback loop-detect, diagnose, correct, verify-is the same loop that operates in monofocal SPC. What EDOF adds is the diagnostic layer between detection and correction. Without that layer, the process engineer knows something is wrong but not what. With it, the process engineer knows what is wrong, where it originates, and how to fix it-all from the same measurement that flagged the issue.
Conclusion
EDOF SPC is monofocal SPC with a diagnostic layer. The fundamental methodology is the same: chart the critical parameters, set control limits from measured data, apply standard rules to detect out-of-control conditions. What EDOF adds is the multi-parameter complexity that requires a hierarchy to manage and a diagnostic framework to interpret.
Three primary charts for real-time monitoring. Four diagnostic charts for root cause identification. The ratio chart as the EDOF-specific innovation that catches what individual charts miss. A three-phase limit calculation approach that starts monitoring on day one and progressively refines the limits as production data accumulates.
The process engineer who sets up EDOF SPC using this framework will detect process drift earlier, diagnose it faster, and correct it with fewer wasted iterations than one who either charts nothing (flying blind) or charts everything (information overload). The hierarchy is the organizational tool that makes twelve parameters manageable and seven control charts actionable.
Monofocal SPC asks one question: is the power right? EDOF SPC asks three: is the plateau right, is the SA recipe right, and is the manufacturing noise low enough to let the recipe work? Seven charts answer all three. The hierarchy tells you which to look at first.
Disclaimer: This document is intended for educational use only. It does not represent legal, regulatory, or certification advice, and should not be interpreted as a declaration of compliance or approval by Rotlex or any regulatory authority. SPC methodologies referenced follow standard industry practice (Montgomery, AIAG). Control limit calculations, rule sets, and parameter selections should be validated for your specific manufacturing process and product.
