Read this first
This is a hypothetical example. The plant, the people and every number below are invented to illustrate how a lean program can unfold. They are not drawn from a real company and should not be used as benchmarks. Real results vary widely, and some plants gain much less.
The imaginary plant
Picture a mid-sized plant that makes machined metal housings for other manufacturers. It has a handful of machines grouped by type: a cutting area, a drilling area, a finishing area and a packing area. Orders come in many varieties and small quantities.
The plant’s problems are familiar. Orders take a long time to get through, there is a lot of unfinished work sitting beside machines, and a noticeable share of parts need rework. Managers feel busy but are unsure why deliveries are late.
For the purposes of this example, the starting figures are:
- Average time from starting an order to shipping it: 9 days.
- Unfinished work in the plant: 1,200 parts.
- Changeovers on the main machine group: 6 per day, 60 minutes each.
- Parts needing rework: 4 percent of 5,000 parts made per day.
All of these are invented.
Step 1: Map the flow
The team follows one product family from raw material to shipping, writing down each step, how long work waits before it, and how long it is actually worked on. They find that most of the 9 days is spent waiting in queues, not being processed. Parts also travel between distant departments, and a batch often waits for the next machine to finish an unrelated job.
This map becomes the shared picture. It also identifies the main bottleneck, a particular machine group where work piles up.
Step 2: Start with the basics
Before changing any machines, the team organizes the bottleneck area. Tools get a marked home, scrap and rework are separated, and the area is cleaned. Operators help decide the layout. These steps cost little and build the habit of looking critically at the work area.
Operators also begin a simple daily check of their own machines, and maintenance staff review the most frequent causes of stoppage.
Step 3: Shorten changeovers
The team records several changeovers on the main machine group and lists every step. They find that much of the 60-minute downtime is spent fetching tools, locating fixtures and adjusting settings by trial and error. By preparing tooling beforehand, marking settings and using quick-release clamps, they work toward a 35-minute changeover. Machine tool setups are reviewed with safety in mind, and isolation procedures stay unchanged.
The arithmetic, using the invented figures:
- Before: 6 changeovers × 60 minutes = 360 minutes per day.
- After: 6 changeovers × 35 minutes = 210 minutes per day.
- Time recovered: 360 − 210 = 150 minutes per day.
The team can use that time to run smaller batches more often, rather than just making more parts.
Step 4: Improve flow and move toward pull
With faster changeovers, batches shrink. The team reworks the equipment layout for the housing family so the cutting, drilling and finishing steps sit closer together, in sequence. Where machines cannot be moved, they set limits on how much work may queue in front of each one, and use simple cards to signal when the next step is ready for more. Work no longer enters the bottleneck area faster than it can leave.
Suppose that after several months, unfinished work falls from 1,200 parts to 700 parts. The reduction is 1,200 − 700 = 500 parts, or about 42 percent (500 ÷ 1,200 is roughly 0.417). Lead time, in this made-up scenario, falls from 9 days to 6.
Step 5: Catch problems sooner
With smaller batches, a problem shows up after a few parts instead of a whole pallet. Operators check critical dimensions at their own station, and the team uses root cause analysis on repeated defects rather than just reworking them.
Imagine rework drops from 4 percent to 3 percent:
- Before: 5,000 × 0.04 = 200 parts per day.
- After: 5,000 × 0.03 = 150 parts per day.
- Difference: 50 fewer parts reworked each day.
Step 6: Look after the machines
Because there is less buffer between machines, a breakdown now stops downstream work almost immediately. The team therefore reviews maintenance priorities. They start planned servicing on the bottleneck machines and begin tracking stoppage causes. If the plant later has enough data, it might consider predictive maintenance for its most critical equipment, but in this scenario it begins with simple routines.
Step 7: Sustain the change
Many programs lose momentum here. The imagined plant takes these steps:
- Written standard methods for the new setups and routines.
- A short daily meeting at the cell, reviewing yesterday’s problems.
- A visible board with a few measures: lead time, defects and breakdown time, in the spirit of common maintenance metrics.
- Ongoing operator training so new staff learn the methods and not the old habits.
- A routine of looking for the next improvement, with ideas from operators encouraged.
What the example shows, and what it does not
The made-up results show how modest changes in several areas can combine: faster changeovers allow smaller batches, smaller batches reduce waiting and speed defect detection, and steadier flow shortens lead time. They also show that many early steps are about organization and observation, not new equipment.
The example leaves out difficulties that real plants often face: skepticism, product mix changes, supplier delays and equipment that does not behave as planned. Gains in practice are often uneven and slower than a tidy example suggests. Any real program should set its own baselines, follow manufacturer guidance and safety procedures, and adapt to local conditions.