General

The Impact of Setup Time

En el corazón de la eficiencia operativa se encuentra una métrica a menudo subestimada: el Tiempo de Preparación o Setup Time. No es simplemente el tiempo que…

By Muestreo del Trabajo ·
The Impact of Setup Time

What Is Setup Time and Why Is It a Critical KPI?

At the heart of operational efficiency lies an often underestimated metric: Setup Time. It is not simply the time an operator takes to change a tool. It is the pulse that sets a plant's agility.

Operationally, we define it as the exact interval between the last good part produced of one batch and the first good part of the next. This concept, popularized by the SMED (Single Minute Exchange of Die) methodology of Shigeo Shingo, is a pillar of modern methods engineering.

Breaking it down is key to improvement. We clearly distinguish between:

  • External Setup: Tasks performed while the machine is still running (preparing tools, documentation). These are eliminable.
  • Internal Setup: Activities requiring full machine stoppage (disassembly, assembly). These are minimizable.
  • Adjustment and Calibration Time: Fine-tuning after the change. Reduced through standardization.

Understanding this taxonomy is not an academic exercise. It is the first step to turning a critical KPI — which can devour up to 35% of productive time — into a competitive advantage. Tools such as Cronometras allow these phases to be broken down and timed with millisecond precision, laying the foundation for any improvement.

The Idle Cost: Real Market Data and Sector Benchmarks

The numbers do not lie. Consolidated Work Sampling studies reveal an impactful reality: in discrete-manufacturing lines with frequent changeovers, between 15% and 25% of total shift time is lost in setup tasks. In high-variability sectors, this figure can rise alarmingly to 35%.

This time is not just an "operational pause". It is an idle cost that directly impacts competitiveness. Below is a sector benchmark based on empirical data:

Sector Average Setup Time (% shift time) Changeovers/Day Negative OEE Impact
Automotive Tier 1 18–22% 4–8 −12 to −18%
Food & Beverage 25–35% 8–15 −15 to −25%
Pharmaceutical 15–20% 2–4 −8 to −15%
Electronics 20–28% 6–12 −14 to −22%
Metal-Mechanical 12–18% 3–6 −10 to −15%
Plastics/Injection 20–30% 5–10 −15 to −22%

Source: Analysis of studies published in indexed industrial-engineering journals.

The table shows a clear pattern: the higher the changeover frequency, the greater the challenge. The food industry, for example, pays a very high toll for its need for flexibility. Diagnosing this problem rigorously is the objective of specialists such as WorkSamp, whose focus on random sampling captures these data without interfering with operations.

Quantifying the Problem: How Setup Time Destroys OEE

For an Operations Director, the universal language is OEE (Overall Equipment Effectiveness). Setup Time directly attacks the most basic factor of this triad: Availability.

The formula is clear:
OEE = Availability × Performance × Quality

Where:
Availability = (Planned Production Time − Stop Time) / Planned Production Time

Setup time is, by definition, a scheduled stop. Let's see its effect with a real numerical example:

  • Shift: 480 minutes.
  • Scheduled breaks: 30 minutes.
  • Net Available Time: 450 minutes.
  • Product changeovers per shift: 5.
  • Current average setup time: 18 minutes per changeover.
  • Total accumulated setup: 90 minutes.

Scenario A (No Setup Management):
Availability = (450 − 90) / 450 = 80%

Scenario B (With SMED Applied):
Setup is reduced to 6 minutes per changeover.
Total accumulated setup: 30 minutes.
Availability = (450 − 30) / 450 = 93.3%

Net gain: +13.3 percentage points in Availability.

This jump is not marginal. Applied to a typical OEE, it could mean moving from 65% to 76% — an increase that translates directly into produced units and freed capacity. Control platforms such as Induly are crucial for real-time monitoring of how these improvements affect total production.

Diagnosis with Statistical Precision: Work Sampling to Measure Setup

This is where classical methods engineering meets modern inferential statistics. We cannot improve what we cannot reliably measure. Continuous timing is invasive and can alter operator behavior (Hawthorne Effect). The solution is Work Sampling based on Tippett's technique.

The foundation is solid: through instantaneous random observations, a "snapshot" of activity at that microsecond is captured. Accumulating hundreds of these snapshots yields a proportion that, thanks to the Central Limit Theorem, approximates a normal distribution (Gauss Curve), enabling inferences with a predefined confidence level.

Calculating sample size (N) is fundamental. We use the formula for a binomial distribution:

N = (Z² × p × (1−p)) / E²

Where:

  • Z: Z value for the confidence level (1.96 for 95%).
  • p: Estimated proportion of setup time (e.g., 0.20).
  • E: Acceptable margin of error (e.g., 0.03).

Practical example:
N = (1.96² × 0.20 × 0.80) / 0.03² = 683 minimum observations.

For statistical robustness, a plant engineer would schedule 700–750 observations randomly distributed across several shifts and days. This method, executed with tools such as WorkSamp, provides a precise, objective diagnosis free from observer bias.

MECE Taxonomy of Setup: Breakdown for Action

Accurate diagnosis requires categorizing all setup time in a MECE way (Mutually Exclusive, Collectively Exhaustive). This taxonomy, applied during the Work Sampling study, identifies exactly where the greatest improvement opportunities lie.

A typical MECE structure for Setup Time would be:

  • S1 — External Preparation (with machine running): Tool search, material transport, blueprint reading.
  • S2 — Disassembly of the Previous System: Removal of tooling, dies, accessories.
  • S3 — Workstation Cleaning: Chip removal, table cleaning, line purging.
  • S4 — Assembly of the New System: Placement and fixing of new tooling.
  • S5 — Initial Adjustment and Calibration: Setting of parameters (pressure, temperature, position).
  • S6 — Verification and Fine Adjustment: First-part production, measurement, corrections.
  • S7 — Waits and Blockages During Setup: Search for support staff, missing crane, missing documentation.
  • O — Other Activities (excludable): Off-topic conversations, unscheduled breaks.

By classifying each random observation into these categories, a Pareto diagram is generated. It is common to discover that S1 (External Preparation) and S7 (Waits) — the easiest to eliminate — represent between 30% and 50% of total setup time. Attacking them first through standardization and organization (5S method) offers quick, visible returns.

Beyond Traditional OEE: Wrench Time and Real Productivity

The impact of Setup Time extends beyond OEE. It affects Wrench Time — the actual time the operator or technician spends working directly on the product or machine. A disorganized, wait-filled setup (category S7) fragments Wrench Time, reducing overall productivity and increasing fatigue.

A well-executed Work Sampling study does not just measure setup. It provides a complete X-ray of the shift:

  • Wrench Time percentage (value added).
  • Setup Time percentage.
  • Waits and Unproductive Movements percentage.
  • Maintenance and administrative activities percentage.

This comprehensive view enables calculating an OEE without hardware sensors on the machine, based purely on statistical observation. It is a non-invasive productivity diagnosis, quick to implement and at a significantly lower cost than full IoT solutions. It is the consulting specialty backed by pure empirical data.

Conclusion: From Measurement to Continuous Improvement

Setup Time is a critical KPI that acts as a lever on industrial productivity. Ignoring it means accepting availability losses of 15–35% — a luxury no competitive operation can afford.

The Work Sampling methodology, grounded in inferential statistics and Tippett's technique, offers the precision diagnosis needed. It allows breaking the problem down with MECE taxonomy, identifying quick wins, and establishing an objective baseline to measure the success of SMED initiatives.

Modern digital tools have democratized access to these methods. From time-analysis platforms such as Cronometras to specialized sampling solutions such as WorkSamp, the path to quantifying and reducing setup is within reach of any plant engineer committed to operational excellence.

The first step, as always, is to measure. Because in industrial engineering, what is not measured with statistical rigor cannot truly be improved.

Resources and Tools