Corvane Medical: Capacity, Changeover, and Postponement
Corvane Medical molds single-use procedure trays on two identical lines at one plant. Each line runs 16 hours a day, five days a week, at a demonstrated 42 trays per line-hour. Weekly demand is 4,600 trays across 14 SKU families that differ only in the final assembly step - handles, labels, and sterile packaging. Changeovers average 3.5 hours and happen 12 times a week; planned maintenance takes 8 line-hours a week. On-time-in-full is 88%, expediting costs $310,000 a year, and finished goods average six weeks of sales at a standard cost of $55. Marketing wants a four-day order-to-ship promise on the top 40% of SKUs. A third-party logistics provider in Columbus has offered to hold unlabeled trays and perform final assembly and labeling there, adding $1.10 per unit on postponed volume, with one-day transit to Corvane's existing customers. Roughly 30% of volume is postponable, unlabeled trays cost $42, and carrying cost is 22% a year.
1. Calculate the plant's realistic capacity and explain why OTIF is 88% when the line appears to have room.
Read the worked answer
Rated capacity: 2 lines × 16 hours × 5 days = 160 line-hours per week. Demonstrated output of 42 trays per line-hour gives a theoretical 6,720 trays.
Realistic capacity: 160 − 42 hours of changeover (12 × 3.5) − 8 hours of maintenance = 110 available line-hours. At 42 trays per hour that is 4,620 trays, against demand of 4,600 (109.5 line-hours).
Utilization is effectively 100% - 109.5 / 110. The plant is not short of annual capacity in aggregate; it is short of protective capacity. Once a line passes roughly 85-90% utilization, queue times and schedule instability rise faster than linearly, so any mix change, quality hold, or unplanned downtime propagates directly into late orders. That is exactly the pattern behind 88% OTIF and $310,000 of expediting, and it explains why adding demand feels impossible even though the arithmetic says there is room.
The gap is variability, addressed three ways:
- Cut changeovers. Dropping from 12 to 8 per week frees 14 line-hours - 588 trays, nearly 13% of weekly demand - without capital. SMED work on the final assembly step is the highest-return project available.
- Add a partial third shift or Saturday overtime as a deliberate capacity buffer sized to variability, not to average demand.
- Decouple the last step, which removes the mix-driven changeover from the constrained molding lines entirely.
2. Evaluate the postponement proposal economically. Should Corvane adopt it?
Read the worked answer
Cost side. 30% of 4,600 = 1,380 trays a week, or about 71,760 a year. At $1.10 that is ≈ $79,000 a year of added conversion and handling.
Benefit side - inventory. Current finished goods: 6 weeks × 4,600 = 27,600 trays at $55 = $1.518M. After postponement, Corvane would hold roughly 2 weeks of standard finished trays (6,440 × $55 = $354,000) plus about 1 week of unlabeled trays at the 3PL (1,380 × $42 = $58,000), totaling about $412,000. The release is $1.106M, worth 22% = ≈ $243,000 a year in carrying cost.
Net: about $164,000 a year, before any credit for reduced markdowns on 14 families, faster response to mix shifts, and reduced expediting.
Break-even test. The carrying benefit divided by postponed volume gives a break-even postponement cost of $243,000 / 71,760 ≈ $3.39 per unit - more than triple the quoted $1.10. The proposal is robust to a poor negotiation with the 3PL, which is the sign of a sound structural change rather than a marginal one.
Caveats to check before signing: sterility and packaging validity once the tray leaves the plant (a regulatory question in medical devices, not just cost); 3PL labeling capacity at peak; whether the freed line time actually materializes as extra molding capacity or is consumed by the same changeover count elsewhere.
Recommendation: adopt, piloted on the five fastest-moving families where mix variance is highest, with contractual quality and capacity terms at the 3PL.
3. Marketing wants four-day order-to-ship on the top 40% of SKUs. What would make that promise credible, and how should the rest of the portfolio be handled?
Read the worked answer
A four-day promise is only credible if it is backed by inventory positioned ahead of the order, not by scheduling intent.
Make it credible:
- The 40% of SKUs should be held as finished or near-finished stock at the 3PL, replenished on forecast, so an order consumes stock rather than triggering a production run. Forty percent of 4,600 is 1,840 trays a week; one week of cover at $42 is about $77,000 of inventory, roughly $17,000 a year of carrying cost - trivial relative to winning the promise.
- Labeling capacity at the 3PL must exceed peak daily demand, not average. If a single day can carry 30% of weekly volume, the 3PL needs capacity for roughly 550 trays a day with a second shift option.
- ATP logic must be truthful. The promise date should be driven by available-to-promise from the 3PL's stock position, with a daily replenishment truck from the plant; a promise the system cannot verify is worse than a longer one.
- Track promise-date adherence as a separate KPI from plant OTIF, because the two failures require different fixes.
The remaining 60%: offer a standard 10-day lead time for configured and slow-moving items, with a small premium option for expedite. Segmenting the promise by demand profile protects the constrained molding lines from low-volume mix churn and keeps changeovers concentrated on families where the setup is genuinely different.
In short, the promise should be a portfolio decision - four days where stock can be positioned, ten days where it cannot - rather than a single service level applied to everything.