Capacity Loss on a Chilled Ready-Meal Line
Northline Foods runs 24 chilled ready-meal SKUs on a single packaging line. The line is scheduled for 120 hours per week and nominally fills 60 packs per minute. Weekly demand averages 280,000 packs with a standard deviation of 22,000. There are 12 planned changeovers per week, each taking 4.0 hours; changeover time is included in the 120 scheduled hours. Overtime is available up to 10 hours per week at a 50% wage premium, but quality rejects rise from 1.8% to 3.5% after the eighth overtime hour. The company promises 98% case-fill to retailers but has averaged 91% for six months, with the worst shortages on the four slowest-moving SKUs. Finished-goods storage is limited to three days. Sales wants more SKUs; operations wants to drop slow movers; finance reports that expediting and markdowns cost ¥18 million last year.
1. Quantify the capacity loss from changeovers and diagnose whether Northline has a capacity problem, a scheduling problem, or both.
Read the worked answer
Gross line capacity = 120 hours × 60 minutes × 60 packs = 432,000 packs per week. Changeover time = 12 × 4.0 = 48 hours, which equals 48 × 60 × 60 = 172,800 packs of lost output. Net capacity = 432,000 − 172,800 = 259,200 packs. Demand is 280,000 packs, so the shortfall is 20,800 packs, or 7.4% of demand. Net capacity covers only 259,200 ÷ 280,000 = 92.6% of average demand before any variability.
This is both a capacity and a scheduling problem. The structural loss is 40% of scheduled hours, which is too high for a 24-SKU line with three-day storage. The 91% service level is consistent with a 7.4% average shortfall plus demand variability. The worst shortages on the four slowest-moving SKUs are a symptom of frequent small batches: slow movers consume disproportionate changeover time and leave little protective capacity for fast movers. Overtime can add 10 × 60 × 60 = 36,000 packs, enough in theory to cover the average gap, but it raises rejects and cost and does not solve the changeover economics.
2. Recommend an operating plan to restore 98% case-fill without a major capital project. Show the logic and the expected capacity effect.
Read the worked answer
The first lever is changeover reduction. If SMED cuts changeover from 4.0 to 2.5 hours, the 12 changeovers consume 30 hours instead of 48. Net capacity becomes 90 hours × 3,600 packs/hour = 324,000 packs, which covers average demand and leaves 44,000 packs of protective capacity. That is a 25% increase in net capacity with no new line.
The second lever is segmentation. Run the four slow movers only in fixed weekly windows or make them to order; do not let them interrupt fast-mover runs. Hold a frozen weekly sequence for the top 20% of SKUs. Use the three-day storage constraint to set a finished-goods policy by SKU: safety stock for fast movers, make-to-order for slow movers, and a service-level agreement that distinguishes retailer-critical SKUs from tail SKUs.
The third lever is overtime discipline. Use overtime only after the eighth hour as a temporary buffer, not as the base plan, because rejects rise from 1.8% to 3.5%. Track changeover hours, schedule adherence, and case-fill by SKU weekly. If changeover cannot reach 2.5 hours within eight weeks, outsource the four slowest SKUs or delist them. The ¥18m expediting and markdown cost is the business case for the change program.
3. Sales insists on adding six new SKUs next year. What conditions should the consultant set before agreeing?
Read the worked answer
Agree only if four conditions are met. First, changeover time must be at or below 2.5 hours, or the line must gain equivalent capacity through outsourcing or a second shift. Second, each new SKU must pass a cost-to-serve test that includes changeover time, storage, markdown risk, and retailer-specific packaging. Third, Northline must adopt SKU rationalization: if six SKUs are added, at least four tail SKUs should be delisted or moved to make-to-order. Fourth, the retailer agreement must accept a segmented service level, not 98% case-fill on every SKU. Without these conditions, adding six SKUs will increase complexity faster than capacity and push service below 91%.