High-Mix Automotive Gear Housing Production in Detroit

Scenario

An automotive transmission plant in Detroit, Michigan, ran mixed-model production of aluminum gear housings (A380 die-cast) on identical Okuma GENOS M460-V CNC mills. With rising labor costs and demand volatility, plant engineers sought to rebalance workloads across three identical cells. A bottleneck had emerged on Cell 2 due to inconsistent cycle times—suspected to stem from unrecorded variations in non-cutting motions during pallet indexing and in-process inspection. They used the CNC Cycle Time Estimator to isolate variables and standardize setup.

Given Data

• Length of cut: 182 mm (average across 4 critical facing/milling operations)
• Feed rate: 0.68 mm/rev (optimized for A380 at 2200 rpm)
• Spindle speed: 2200 rev/min
• Time for tool changes: 0.33 min (automated tool changer, well-maintained)
• Non-cutting motions time: initially estimated at 0.45 min—but stopwatch data revealed actual average was 0.91 min due to redundant pallet rotation and manual gage verification.

Calculation

Cutting Time = 182 mm ÷ (0.68 mm/rev × 2200 rev/min)
= 182 ÷ 1496 ≈ 0.1217 min → 0.12 min

Total Cycle Time (initial estimate) = 0.12 + 0.33 + 0.45 = 0.90 min
Total Cycle Time (actual) = 0.12 + 0.33 + 0.91 = 1.36 min

The 0.46-min discrepancy explained the observed 51% throughput gap between Cell 2 and Cells 1 & 3.

Result and Decision

Engineers reprogrammed the pallet indexing sequence to eliminate one full rotation and integrated automated vision-based in-process inspection—reducing non-cutting motions time to 0.52 min. Recalculating: Total Cycle Time = 0.12 + 0.33 + 0.52 = 0.97 min, aligning closely with Cells 1 & 3 (0.95–0.98 min). Cell 2 throughput increased by 40%, eliminating the need for costly overtime.

Lesson

Non-cutting motions time is often the largest source of hidden cycle time variance in high-mix environments—never rely on default estimates; measure it empirically under real production conditions before optimizing cutting parameters.