Case Study: Manufacturing Throughput Crisis

Industry: Precision Manufacturing Problem Type: Operations Duration: 6 weeks Outcome: 45% throughput increase

The Situation

A precision manufacturing company producing aerospace components faced a throughput crisis. Despite significant investment in new CNC equipment, output had plateaued. Customer orders were backlogging, and the company was at risk of losing key contracts.

What we heard initially:

• "The new machines aren't performing to spec"
• "We need to add a night shift"
• "The operators need more training"

The Investigation

Week 1-2: Symptom Mapping

We started by documenting exactly what "throughput crisis" meant:

• Output: 850 units/week (target: 1,200)
• Utilization: 62% (claimed), actual TBD
• Backlog: 6 weeks and growing
• Quality: 4.2% rejection rate

Week 2-3: Process Tracing

We followed the flow from raw material to shipping:

1. Material receiving: No delays observed
2. Machining (3 CNC cells): High variability in cycle times
3. Deburring: Manual, consistent pace
4. Inspection: Significant queuing observed
5. Assembly: Waiting for inspected parts
6. Shipping: On-time once parts available

The inspection queue was the first clue. Parts were piling up waiting for quality verification.

Week 3-4: Root Cause Analysis

Inspection Bottleneck

The quality department had three inspectors. Inspection time averaged 12 minutes per part. Capacity: ~120 parts per day across all inspectors.

Production was capable of ~200 parts per day.

No matter how fast production ran, inspection could only process 120 parts daily. The gap accumulated as backlog.

But why was inspection so slow?

We dug deeper. Inspectors were following an outdated procedure that required:

• 23-point manual measurement
• Full documentation on paper forms
• Secondary verification for every part

This procedure was written when the company produced 15 different part numbers. They now produced 180. The procedure had never been updated.

Secondary finding: Setup time waste

While investigating, we noticed machine setup times varied wildly: 45 minutes to 4 hours for the same part. Tribal knowledge problem. Some operators knew the tricks; others didn't. See Field Note: Tribal Knowledge.

The Findings

Root cause 1: Inspection procedure misaligned with current production volume and part variety. Designed for a different era.

Root cause 2: Undocumented setup procedures leading to variable machine utilization.

Contributing factors:

• No capacity planning across departments
• Inspection treated as separate from production
• Resistance to procedure changes due to aerospace certification concerns

The Recommendations

Immediate (Week 1-2)

1. Risk-stratify inspection

   • Not all parts need 23 measurement points
   • Implement tiered inspection based on part criticality and history
   • Predicted reduction: 23 points → average of 8 points

2. Cross-train for flexibility

   • Train two production supervisors on basic inspection
   • Allow overflow capacity during peaks

Short-term (Month 1-2)

3. Document setup procedures

   • Capture best practices from experienced operators
   • Create standardized setup sheets per part family
   • Target: 90-minute maximum setup time

4. Digital inspection forms

   • Eliminate paper documentation
   • Reduce transcription errors
   • Enable real-time quality tracking

Medium-term (Month 2-4)

5. Capacity alignment

   • Match production scheduling to inspection capacity
   • Eliminate production of parts that will sit in queue

6. Quality engineering review

   • With aerospace auditor present
   • Formally revise inspection procedures
   • Certify the streamlined process

The Implementation

The client implemented recommendations 1-4. Recommendation 5 was partially implemented. Recommendation 6 was deferred due to an upcoming customer audit.

The Outcome

After 3 months:

• Output: 1,240 units/week (target exceeded)
• Utilization: 78%
• Backlog: 2 weeks (within normal)
• Quality: 3.8% rejection rate (improved)

Quantified impact:

• 45% throughput increase
• $2.1M additional annual revenue capacity
• 0 new equipment purchases needed
• 0 additional headcount required

Key Lessons

The bottleneck was invisible

Everyone was looking at production. The constraint was in quality. See Field Note: Finding the Bottleneck.

Procedures fossilize

What was appropriate 10 years ago wasn't appropriate today. But no one had questioned it. "That's how we've always done it."

Documentation matters

The setup time variability was pure tribal knowledge. Once captured, the benefit was immediate.

Systems, not people

The inspectors weren't slow. The procedure was wrong. The operators weren't inconsistent. The knowledge transfer was missing.

The solution to a production problem was found in the quality lab.