Introduction: The Production Reality – One Floor, Four Products
The textbook IOL manufacturing line produces one product. Monofocal IOLs flow through the line at steady cadence. The acceptance criteria are stable. The operators perform a single sequence of measurements thousands of times per week. The SPC charts are simple. The throughput is predictable.
Few facilities have this luxury. The commercial reality is that a single production floor produces four or more IOL categories: standard monofocal, toric monofocal, multifocal, and EDOF-often with hydrophobic and hydrophilic variants of each, in multiple powers, with optional add powers and toric corrections. The product matrix can easily reach 60–80 distinct SKUs running on the same floor, the same equipment, and largely the same operators.
This multi-product reality creates challenges that single-product engineering does not address. Changeover time between products consumes 8–16 hours per week of effective production capacity. Operator cognitive load increases as the protocol switches between product types. Acceptance criteria, SPC charts, and reference lenses multiply with each product family. The probability of cross-product errors-applying monofocal criteria to an EDOF lens, or vice versa-grows with the number of product transitions per shift.
The challenges are not insurmountable. They are addressable through specific design decisions about floor layout, workflow architecture, changeover protocols, operator training, and data infrastructure. This article describes the practical engineering of a multi-product IOL manufacturing floor: what to colocate, what to separate, how to sequence product transitions, and how to architect the QC system so that running four products on one floor does not produce four times the errors.
Why Multi-Product Manufacturing Creates Non-Linear Complexity
If running monofocal production at 100% efficiency consumes 100 hours of equipment time per week, naive planning would expect that running four products in equal proportions would consume the same 100 hours-just divided into 25-hour blocks. This linear assumption is wrong in three specific ways.
Changeover time is non-productive time
Each product transition requires changeover activities: loading new acceptance criteria, swapping reference lenses, recalibrating verification gauges, briefing operators on the new product, and verifying the system before resuming production. A typical product changeover consumes 90–180 minutes of equipment and operator time during which no production lenses are inspected.
With four products and one transition per day, weekly changeover time is 5 transitions × 120 minutes = 600 minutes = 10 hours. This is 10% of a 100-hour production week consumed by non-productive changeover. With more frequent transitions (e.g., two products per shift to match a complex product mix), the changeover overhead can reach 16–20% of weekly capacity.
Cognitive load increases error rates
An operator running monofocal-only inspection executes the same protocol thousands of times. The procedure becomes automatic. Errors are rare because the procedure is highly practiced.
An operator running four products in alternating batches must continuously context-switch. After 5 hours of monofocal inspection, the operator transitions to EDOF. The new protocol requires through-focus interpretation-a different cognitive task. The operator’s familiarity with the EDOF protocol is lower because they performed it less frequently. The probability of error increases-typically 2–3× the error rate of a single-product workflow.
Verification overhead multiplies
Each product requires its own reference lens for system verification, its own acceptance criteria for product code loading, and its own daily verification protocol. A monofocal-only line verifies one reference lens at the start of each shift. A four-product line verifies four-or, more practically, verifies the reference lens for the next product immediately before each changeover.
The verification overhead is multiplicative: 4 products × daily verification = 4 verification events per day, each consuming 15–20 minutes. The EDOF system verification protocol, in particular, requires an EDOF reference lens that does not interchangeably serve monofocal verification-adding a verification step that single-product facilities do not have.
The Product Matrix: What Each IOL Type Requires
Multi-product floor design starts with a clear understanding of what each product requires from the production and QC systems. Different products have different measurement protocols, different acceptance criteria, different operator skills, and different throughput characteristics.
Table 1: Per-Product Requirements Matrix
| Product | Material | Measurement Mode | Batch Size | Cycle Time | Operator Skill Required |
| Standard monofocal | Hydrophobic / Hydrophilic | Power, cylinder, axis | 50 dry / 12 wet per cycle (IOLA MP) | 4 sec/lens dry; ~50 sec batch | Standard. Power threshold disposition. Single-shift training. |
| Toric monofocal | Hydrophobic / Hydrophilic | Power + cylinder + axis verification + alignment | 50 dry / 12 wet (IOLA MP) | 4 sec/lens (axis automatic) | Standard + axis interpretation. Toric reference lens for daily verification. |
| Multifocal (diffractive) | Mostly hydrophobic | Through-focus discrete peaks + add power verification | Sampled lenses on IOLA MFD; 100% on IOLA MP | 9 sec/lens MFD; 4 sec MP | Through-focus peak interpretation. Add power threshold. Distinct from EDOF protocol. |
| EDOF | Hydrophobic / Hydrophilic | Through-focus plateau + multi-aperture + SA coefficients | Sampled on IOLA MFD; 100% on IOLA MP | 9 sec/lens MFD; 4 sec MP | Plateau interpretation (different from multifocal peaks). Four-outcome disposition. Tightest cylinder spec. |
| Toric EDOF | Mostly hydrophobic | Through-focus plateau + axis + multi-aperture | Sampled on MFD; 100% on MP | 9 sec/lens MFD | Most complex protocol. Combines EDOF and toric requirements. Requires highest operator skill. |
| Toric multifocal | Hydrophobic | Through-focus peaks + axis + add power | Sampled on MFD | 9 sec/lens MFD | Multifocal + toric. Add power verification. Distinct acceptance criteria from EDOF. |
The matrix reveals several practical insights. First, the same hardware (IOLA MP for batch and IOLA MFD for through-focus) handles all products-the constraint is software configuration, not equipment. Second, the throughput characteristics are similar within categories (4 seconds per dry lens, 9 seconds for through-focus) but the protocol complexity differs significantly. Third, hydrophilic variants of any product require wet measurement workflows that change the batch size and the hydration management overhead.
Floor Layout Architectures: Three Options
How the production floor is physically organized determines how easily multi-product workflow operates. Three architectural patterns are common, each with distinct tradeoffs.
Architecture A: Single shared QC station, sequential batches
All products flow through a single QC station equipped with the IOLA MP for 100% inspection and the IOLA MFD for through-focus verification. Batches are scheduled sequentially: monofocal Monday morning, toric Monday afternoon, EDOF Tuesday morning, multifocal Tuesday afternoon. The station performs one product at a time.
Pros: Lowest equipment cost. Single instrument set. Simplest data flow. Smallest physical footprint. Lowest operator headcount.
Cons: Highest changeover overhead because every product transition requires full system reconfiguration. Production capacity is the most constrained. Bottleneck at the QC station limits the entire upstream production rate.
Best for: Smaller facilities (under 5,000 lenses per week total) or facilities just transitioning into premium IOL production where premium volume is still building.
Architecture B: Parallel QC stations by product family
Two QC stations run in parallel: one dedicated to standard products (monofocal, toric monofocal) using primarily the IOLA MP, the other dedicated to premium products (multifocal, EDOF, their toric variants) using both the IOLA MP and IOLA MFD. Products are routed to the appropriate station based on category.
Pros: Reduced changeover within each station because each station handles fewer product types. Premium and standard products can run simultaneously. Operators specialize: standard-station operators master monofocal/toric protocols deeply; premium-station operators master through-focus interpretation. Higher cumulative throughput.
Cons: Higher equipment cost (two stations). Larger floor footprint. More operators. Asymmetric utilization-premium volume may not yet justify dedicated equipment if premium fraction is below 25–30%.
Best for: Mid-size facilities (5,000–20,000 lenses per week) with established premium volume above 25%. The dominant architecture for facilities scaling EDOF production.
Architecture C: Multiple parallel stations with full flexibility
Three or more QC stations operate in parallel, each capable of any product. Stations are not dedicated to product families-they are dynamically assigned based on production scheduling. A station running EDOF in the morning may switch to toric in the afternoon based on the production plan.
Pros: Maximum throughput flexibility. Can absorb production spikes in any product category. Minimal cross-station coordination because every station has full capability. Easy to scale by adding stations.
Cons: Highest equipment cost (3+ full stations). Highest operator skill requirement (every operator must master every product). Most complex training and certification matrix. Largest floor footprint.
Best for: Large facilities (above 20,000 lenses per week) with diverse premium portfolio and significant volume in each product category. Justified when individual product volumes are high enough that dedicated single-product stations would be underutilized.
Table 2: Floor Layout Architecture Comparison
| Dimension | A: Single Station | B: Parallel by Family | C: Full Flexibility | Driving Factor | Recommendation |
| Capital cost (QC equipment) | Lowest (1 × IOLA MP + 1 × MFD) | Medium (2 × MP + 1–2 × MFD) | Highest (3+ MP + 2+ MFD) | Total volume + premium fraction | Match to current and 3-year forecast volume |
| Weekly capacity (lenses/week) | ≤ 5,000 | 5,000–20,000 | > 20,000 | Aggregate production volume | Plan for next architecture before reaching capacity |
| Changeover overhead | 16–20% of capacity | 8–12% of capacity | 4–8% of capacity | Number of product transitions | Architecture B is the typical inflection point |
| Operator headcount | 1–2 per shift | 3–4 per shift | 5+ per shift | Number of stations + redundancy | Cross-train all operators on standard products |
| Operator skill complexity | Highest per operator (knows all products) | Specialized (standard or premium) | Highest per operator (full flexibility) | Training cost + retention risk | Specialization (B) reduces cognitive load |
| Cross-product error risk | Highest (frequent transitions per operator) | Lowest (specialized operators) | Medium (full flexibility but smaller batches per product per operator) | Quality risk + complaint cost | Architecture B minimizes error risk through specialization |
Changeover Optimization: Where the Hours Disappear
Changeover time is the largest single opportunity to reclaim productive capacity in multi-product manufacturing. Reducing changeover from 120 minutes to 60 minutes across 5 weekly transitions reclaims 5 hours of weekly production capacity-equivalent to 75 additional EDOF lenses per week at 4 minutes per lens including handling.
The changeover task list
A typical product-to-product changeover at a QC station includes the following activities: (1) close out current product batch records and verify all measurements complete, (2) load new product code and acceptance criteria into the measurement system, (3) swap or verify reference lens for new product, (4) perform system verification with new reference lens, (5) brief operator on new product specifics (if shift handover involved), (6) load new product batch and resume production.
Of these six activities, only steps 1 and 6 are unavoidable. Steps 2–5 are candidates for optimization through configuration management, parallel execution, and pre-staging.
Configuration management: product codes
Modern measurement systems support product code libraries: each product is pre-configured with its acceptance criteria, measurement modes, and data export rules. Loading a product code is a single barcode scan that takes seconds. The configuration overhead is moved upstream into the product setup phase, performed once per product family rather than once per changeover.
This requires discipline in product code management. Every SKU has a documented product code. Every acceptance criteria revision triggers a controlled product code update. The library grows over time but remains manageable through systematic naming and version control.
Reference lens management
Reference lenses for system verification should be pre-staged in clearly labeled storage at each QC station. The EDOF reference lens for system verification is distinct from the monofocal reference and from the multifocal reference. Storing all references at the station-with the active reference for the current product clearly identified-eliminates the search time that often consumes 5–10 minutes of every changeover.
Parallel execution
System verification (Step 4) can run in parallel with batch closeout (Step 1) when staffing permits. While one operator finalizes the previous batch records, another loads the new product code and runs the verification reference lens. The two activities take place simultaneously instead of sequentially. With proper choreography, this parallelization reduces total changeover time by 20–40%.
Pre-staging at shift boundaries
If a product transition is scheduled across a shift change, the pre-staging happens during the shift overlap rather than after the previous shift ends. The arriving shift loads product codes, prepares reference lenses, and reviews the new product’s SOP-all before the previous shift’s production work is complete. The transition is effectively complete before the new shift’s production starts.
Operator Skill Matrix and Training
Operator skill is a critical resource in multi-product manufacturing. Different products require different cognitive frameworks for result interpretation-a fact that simple training plans often underestimate.
The skill matrix
Table 3: Operator Skill Matrix – Product vs Required Capability
| Capability | Standard Monofocal | Toric Monofocal | Multifocal / EDOF |
| Power threshold disposition | Required | Required | Required (still part of EDOF) |
| Cylinder threshold disposition | Required (loose tolerance) | Required (label match) | Required (tight tolerance for non-toric EDOF) |
| Axis interpretation | Not required | Required (label match) | Required for toric variants only |
| Through-focus peak recognition (multifocal) | Not required | Not required | Required (multifocal): discrete peaks normal; valleys normal between peaks |
| Through-focus plateau recognition (EDOF) | Not required | Not required | Required (EDOF): continuous plateau normal; valleys are defects |
| Add power verification (multifocal) | Not required | Not required | Required for multifocal; not applicable for EDOF |
| Plateau width / minimum MTF threshold (EDOF) | Not required | Not required | Required for EDOF; not applicable for multifocal |
| Four-outcome disposition (EDOF) | Not required | Not required | Required for EDOF: pass/fail-power/fail-plateau/both-fail logic |
| Multi-aperture verification (EDOF) | Not required | Not required | Required for sampled EDOF lenses |
Training cadence
Training new operators on a multi-product floor follows a tiered approach. The first 4–6 weeks focus on standard monofocal: power, cylinder, threshold disposition, daily verification. The operator becomes proficient on the simplest protocol before more complex products are introduced.
Weeks 6–10 add toric monofocal: axis interpretation, label match verification, toric reference lens. The operator now handles two product types reliably.
Weeks 10–16 add multifocal or EDOF (whichever is more central to the product mix): through-focus interpretation, the appropriate disposition logic, and the product-specific reference lens. Pattern recognition training-distinguishing passing from failing through-focus profiles-is the largest skill addition.
Weeks 16+ add the second through-focus product (the one not added in weeks 10–16). The critical training point is distinguishing the two: multifocal valleys are normal; EDOF valleys are defects. This distinction must be reinforced repeatedly because operators who learn one product first often default to that interpretive framework when measuring the other.
Total time to a fully cross-trained operator: 4–5 months. This is significantly longer than single-product training (1–2 months) and represents a real cost. The investment is worthwhile for facilities operating Architecture C (full flexibility), where every operator must handle every product. For Architecture B (parallel by family), operators specialize and the training is shorter.
Data Architecture: Multi-Product SPC and Traceability
Multi-product manufacturing multiplies the data complexity. Each product generates its own measurement records, its own SPC charts, its own batch dispositions, and its own complaint history. The data infrastructure must accommodate this multiplicity without becoming unmanageable.
Product-specific SPC charts
SPC charts are product-specific because the control limits, target values, and parameter sets differ. The monofocal sphere power chart has different limits than the EDOF sphere power chart-because the underlying production processes have different distributions.
When the operator views the SPC dashboard, the charts displayed should match the product currently in production. Showing all products’ charts simultaneously creates information overload. Showing only the current product’s charts ensures the relevant control limits are in view.
Cross-product trend visibility
While individual product SPC is product-specific, certain trends are visible across products. If parasitic RMS increases simultaneously across all products, the source is likely environmental or equipment-related rather than product-specific. A separate cross-product summary view-typically at the supervisor level rather than the operator level-provides this visibility for diagnostics.
Batch traceability
Multi-product batches must be traceable backward through the production history. A complaint on an EDOF lens implanted six months ago should retrieve: the production batch, the operators on shift during measurement, the system verification record from the start of that shift, the SPC state at the time of measurement, and the complete measurement data for that specific lens.
This traceability is straightforward in single-product manufacturing because every record is tagged with the product. In multi-product manufacturing, the tagging discipline is more important and more complex. Every measurement record must include product code, batch number, operator ID, and timestamp. Every system verification record must include the product context. Every SPC chart must be filterable by product.
Practical Example: A 5,000-Lens-Per-Week Mixed Mix
To make the architecture decisions concrete, consider a representative facility producing 5,000 lenses per week across the following mix:
- 2,500 standard monofocal (50%)
- 1,000 toric monofocal (20%)
- 800 EDOF (16%)
- 500 multifocal (10%)
- 200 toric EDOF / toric multifocal combined (4%)
Total premium fraction: 30%. This is a typical facility transitioning from primarily monofocal to substantial premium volume.
At an IOLA MP throughput of 4 seconds per lens (50 lenses per 200-second cycle), 5,000 lenses requires 200 cycles or approximately 11 hours of pure measurement time per week. Plus changeovers (estimated 8 hours weekly at this volume) and verification (4 hours weekly), the total QC equipment time is approximately 23 hours weekly. A single station running one shift covers this comfortably.
However, the through-focus verification on EDOF and multifocal samples (10 samples per batch × 1 batch per day per product = 30 lenses per day on the IOLA MFD = 270 seconds per day) is small but cannot share equipment with the IOLA MP because both may need to operate on the same batch.
Architecture B (parallel by family) is the appropriate choice. Station 1 (IOLA MP) handles all batch inspection. Station 2 (IOLA MFD) handles through-focus sampling for EDOF and multifocal. Operators specialize: Station 1 operators master batch protocols; Station 2 operators master through-focus interpretation.
As volume grows toward 10,000+ lenses per week, the facility may add a second IOLA MP for redundancy and capacity expansion. As premium fraction grows above 35–40%, a second IOLA MFD or a dedicated EDOF-only station may be justified.
The architecture is not static. It evolves with the product mix and the volume. The scaling considerations for this evolution are addressed in the companion article on scaling premium IOL production.
Common Multi-Product Mistakes and How to Avoid Them
Mistake 1: Treating EDOF as “another monofocal”
The most common multi-product error is operating EDOF inspection with monofocal-era protocols. Power and cylinder are measured. Through-focus is not. The lens passes the monofocal test and ships without verification of the parameter that defines its premium value. The hidden cost of this gap appears 3–6 months later as field complaints with no diagnostic data. Solution: ensure EDOF product codes mandate through-focus verification, and verify this configuration daily.
Mistake 2: Single reference lens for all products
Using a monofocal reference lens to verify the system before EDOF production verifies power accuracy but not through-focus computation. Solution: a separate reference lens for each product family, with clear SOP requirements for which reference is used for which product.
Mistake 3: Identical disposition logic across products
A binary pass/fail disposition works for monofocal. EDOF requires the four-outcome disposition (full pass, power-pass-plateau-fail, power-fail-plateau-pass, both fail). Forcing EDOF results into a binary system loses the diagnostic information about which parameter failed. Solution: product-specific disposition workflows in the QC software and operator training.
Mistake 4: Ignoring material variant overhead
Hydrophilic variants of any product require wet measurement protocols with smaller batch sizes (12 lenses per cycle vs 50 dry) and hydration equilibrium time (24–48 hours). Adding hydrophilic variants can quadruple the equipment time per lens. Solution: schedule hydrophilic batches separately, plan instrument utilization based on the wet-versus-dry mix, and account for hydration staging space.
Mistake 5: Operator rotation without re-certification
Operators rotated from one product family to another after a long absence may need recertification, particularly for through-focus interpretation skills. Solution: establish a competency refresh requirement (e.g., monthly verification of through-focus pattern recognition) for operators who do not perform a product type at minimum quarterly.
Conclusion
Multi-product IOL manufacturing is the commercial reality. Few facilities can afford to dedicate production lines to single products, and most operate four or more product categories on shared equipment, shared operators, and shared data systems. The challenge is not whether to run multiple products but how to do so without compounding the complexity.
The hardware does not constrain multi-product manufacturing. The same IOLA MP measures monofocal, toric, multifocal, and EDOF with the right product code loaded. The same IOLA MFD captures through-focus data for any presbyopia-correcting design. Equipment capacity is not the limiting factor.
The constraints are workflow design, operator cognitive load, changeover overhead, and data architecture. Architecture B-parallel QC stations specialized by product family-addresses these constraints for the majority of mid-size facilities. It reduces changeover within stations, allows operator specialization, and scales naturally as premium volume grows. Architecture A is appropriate for smaller facilities; Architecture C for large facilities with diverse high-volume portfolios.
The optimization opportunities are concrete. Configuration management with product code libraries reduces changeover by 30–50%. Parallel execution of verification and batch closeout reduces it further. Pre-staged reference lenses eliminate search time. Product-specific SPC charts prevent operator information overload while preserving cross-product trend visibility for supervisors. Operator training paths from monofocal to through-focus build skill in the order that minimizes cognitive load.
One floor produces four products. The hardware does not care. The operators do, the workflow does, and the data architecture does. Designing the floor for multi-product reality-rather than retrofitting a monofocal layout to handle premium volumes-is the difference between absorbing the multi-product overhead and being absorbed by it.
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. Architecture recommendations and overhead estimates are illustrative and depend on specific facility characteristics, product mix, and operational practices.
