Live data flowing in 7 days. No IT department required. See exactly what it looks like for your machines.
Calculate Your Downtime CostPractical guides, benchmarks, and shop floor strategies — no fluff, no spam, unsubscribe anytime.
The TPM Six Big Losses framework applied to CNC machining. Equipment failure, setup time, micro-stops, reduced speed, defects, and startup losses — with data on where most shops lose the most.
Reactive vs preventive vs predictive maintenance for CNC shops. What signals to monitor, how vibration analysis works, real-world savings for a 5-machine shop, and common pitfalls to avoid.
Lights-out machining adds 50-75% capacity at near-zero labor cost. But you cannot run unattended without monitoring. Here is what to monitor, how to build confidence in stages, and the ROI math.
Somebody told you “world-class OEE is 85%.” That number comes from Seiichi Nakajima's TPM research in the 1980s, and it was derived from high-volume automotive production lines running the same part 24 hours a day. If you run a job shop with 50 different part numbers per week, that benchmark is worse than useless — it is demoralizing.
This article provides CNC-specific OEE benchmarks broken down by machine type and production style. Not aspirational targets from consultants — realistic percentiles based on what shops actually achieve.
The 85% world-class number persists because it is simple and quotable. But it collapses three very different factors — Availability, Performance, and Quality — into one number, and it does not account for production style at all.
Here is why the number is misleading for CNC shops:
These benchmarks reflect high-mix, low-volume job shop environments. Dedicated production cells running the same part will trend 5-10 points higher across the board.
| Machine Type | 50th Percentile | 75th Percentile | 90th Percentile | Primary Loss Driver |
|---|---|---|---|---|
| 3-Axis Vertical Mill | 45-52% | 65-72% | 80-85% | Setup / changeover time |
| CNC Turning Center | 50-58% | 68-75% | 82-88% | Tool wear / insert changes |
| 5-Axis Mill | 40-48% | 60-68% | 75-82% | Programming / prove-out time |
| Swiss-Type Lathe | 55-62% | 72-78% | 85-90% | Bar stock changes |
| Horizontal Machining Center | 50-58% | 70-76% | 82-88% | Pallet change / fixture setup |
| Multi-Spindle Screw Machine | 58-65% | 75-82% | 85-92% | Spindle synchronization |
Percentiles based on single-shift, 5-day operation with planned maintenance excluded from production time. Shops running 2+ shifts typically see 3-5 points higher OEE due to amortized setup time.
The OEE number is the product of three factors. Understanding where your points go missing matters more than the final number. Here is the typical breakdown for a CNC job shop at the 50th percentile:
| OEE Factor | 50th Percentile | 75th Percentile | Key Loss Sources | Where to Look First |
|---|---|---|---|---|
| Availability | 70-78% | 85-90% | Changeovers, breakdowns, waiting | Setup time tracking |
| Performance | 75-82% | 88-93% | Feed overrides, micro-stops, slow cycles | Cycle time vs. ideal |
| Quality | 95-98% | 98-99.5% | Scrap, rework, first-article rejects | Reject tracking by job |
The math reveals a pattern: most CNC shops have decent Quality (95%+) but lose heavily on Availability and Performance. A shop at 75% Availability, 80% Performance, and 97% Quality has an OEE of 58%. That is right at the 50th percentile for a vertical mill — and it is not a Quality problem.
This is why generic advice to “improve quality” misses the mark for CNC. The biggest OEE gains in machining come from reducing setup time (Availability) and eliminating micro-stops (Performance). Quality improvement matters, but it is typically the third priority, not the first.
Not every shop needs to chase 85%. The right OEE target depends on your business model:
The point is not to hit an arbitrary number. The point is to know your current number, understand where the losses are, and make deliberate decisions about which ones to attack. A 15-point improvement from your current baseline — wherever that baseline is — typically translates to $225,000 in recovered capacity across 5 machines.
The number one reason shops have unreliable OEE numbers is manual data collection. Operator-reported times are consistently optimistic — setup times are underreported by 20-40%, micro-stops go unrecorded, and slow cycles are invisible.
Accurate OEE measurement requires:
Compare your OEE anonymously
Our free benchmark tool calculates your estimated OEE by machine type and production style. See how your shop compares to similar operations. No login required.
Run a free OEE benchmark for your shop →The real question is: “Where are we losing the most, and what would it cost to fix it?”
A shop at 50% OEE that improves to 65% across 5 machines running at $150/hr has recovered $225,000 in annual capacity. That is not a theoretical number — it is 1,500 additional productive hours that were already paid for in labor, rent, and machine depreciation.
The monitoring that makes this visible costs $4,788 per year. The improvement it enables is 47x that cost. Even capturing a quarter of the available improvement — a realistic first-year target — yields a 12x return.
Stop comparing your CNC shop to a Toyota assembly line. The 85% world-class benchmark was not designed for your production environment. CNC job shops at the 50th percentile run 45-55% OEE, and getting to the 75th percentile (65-72%) is a meaningful, achievable goal that translates directly to revenue.
Measure your actual number. Break it down by Availability, Performance, and Quality. Attack the biggest loss first. Repeat.
The goal is not a perfect score. The goal is knowing your score and making it better every month.