Hourly production counts: the simplest way to cut rejections

Every factory has rejections. The expensive ones are not the rejections themselves, but the ones you find late: a whole lot of handles with holes drilled off-centre, discovered at buffing, after five more stages of work have gone into them.

The fix is surprisingly simple: count rejected pieces at every stage, every hour, and look at the numbers the same day. Here is why it works and how to set it up.

Why rejections found late cost the most

A rejected piece carries the cost of every stage it went through. A piece rejected at the first stage wastes a little raw material. The same defect caught at the last stage wastes the material, plus the labour and machine time of every stage in between.

When rejections are only counted at final inspection, or totalled at month end, you lose two things:

• Where it happened. A pile of rejects at the end doesn't tell you which stage caused them.
• When it happened. By the time you notice, the tool has been worn or the setting has been wrong for days.

What hourly counts change

When each station records good and rejected pieces every hour:

• A spike shows up within the hour, at the stage that caused it.
• The supervisor can check the tool, the setting or the material before the next hundred pieces are made.
• You can compare stages and products: which stage rejects the most, and is it getting better or worse?

You also get a fair picture of each stage. A station with low output might simply be catching problems that earlier stages let through.

How to set it up

1. Agree on what counts as a reject at each stage. Write it down and put it at the station. "Hole more than 1 mm off-centre" is better than "bad hole".
2. Keep rejected pieces in a separate, marked bin at every station, so the count is easy and honest.
3. Record two numbers per lot every hour: good and rejected.
4. Look at the rejection percentage daily: rejected ÷ (good + rejected). Watch the trend, not just the number.
5. Talk about the top stage every morning. Five minutes on the stage with the most rejections yesterday is enough to start.

Make it easy to tell the truth

Counts are only useful if operators record rejections honestly. Make it clear that recording a reject is good work: it stops a bad piece from wasting more time downstream. Never punish someone for the number they write. Use it to fix tools, settings and training.

How LineSight does it

On the LineSight entry screen, every lot at a station has two fields: Good and Rejected. The dashboard shows rejected pieces today for each stage, highlighted in red, along with the factory's rejection percentage for the day. Each lot keeps a full history of who entered what and when, and the daily output report breaks down good and rejected pieces by product and stage, ready to download as a CSV.

The takeaway

Rejections are cheapest when they are caught early and close to their cause. Count good and rejected pieces every hour at every stage, look at the numbers daily, and focus on the worst stage first. It costs almost nothing, and it is one of the fastest ways to cut waste on any shop floor.