Kaizen: shop floor improvements that stay in the standard.
Kaizen in manufacturing is a team improvement in one area of the plant: over a few days the team removes the waste it sees in the work and writes the new way of working down as a standard. With SolveR the standard leaves the event with an owner and a review date, and the review records whether the gain held.
Kaizen in SolveR is set up by agreement: on the discovery call we show how it would run on your case and tell you when you can start using it.
When to use kaizen, and when not to
Kaizen pays off when one area has several kinds of waste at once, such as waiting and walking, and they have to be seen where they happen. A team of people from the area and from outside it works on the spot for a few days, and an event leader makes sure the actions actually get closed. For a single problem with an obvious fix it is too heavy.
- Just-Do-It for the obvious fix at one workstation, done right away and written down at once.
- PDCA when you have one idea and want to test it.
- DMAIC when a deviation keeps coming back, the cause is unknown and the data has to show it.
All methods in one place, with their typical duration: How it Works.
A press changeover, today and with SolveR
Typical case: a kaizen event on a press where the tool changeover takes too long.
| Today | With SolveR |
|---|---|
| The charter, observations and fishbone are on paper, and the report-out for management is rebuilt in PowerPoint on the last day. | The purpose, target and team are recorded before the event, and the baseline has a source: a time study of the changeover. |
| The observation "the operator walks a lot" stays without a quantity, and the type of waste is picked by feel. | Every observation from the walk has a type of waste and a quantity: trips for screws and keys, centring the tool by hand, the press standing idle while the tool heats up. |
| An action is ticked off and a trial is called a success, although the target was not met. | The cause is confirmed by the time study: much of the changeover can be done while the press is still running. The actions are a cart with a prepared kit and a standard sequence of steps, and the trials show whether the target is met. |
| After the event, actions are left without an owner, the review is skipped and the next event starts from a blank flip chart. | The standard has an owner and a review date, the other shifts are trained, and the review records whether the gain held. |
The case is an illustration written for this page.
What a kaizen event leaves behind
A one-page record of the event: purpose and target, observations with their type of waste, the confirmed cause, actions with their trials, and the new standard. The standard has an owner and a review date, so everyone knows who holds it and when someone checks whether the gain held. Because the observations, cause and actions are linked, the record shows why the standard is the way it is.
The sponsor who agreed the scope and the target reads one page and sees what changed in the area.
The steps of a kaizen event
- PreparePurpose, target, team, a baseline with a source
- ObserveA walk on the spot, observations tagged by waste
- AnalyseA cause confirmed by measurement
- ImproveActions and trials in the area
- StandardiseThe new way of working, an owner, a review date
The event in the area takes a few days, and preparation starts a few weeks before.
One knowledge base for every method
A closed kaizen event stays in the knowledge base with its observations, cause, actions and standard. When the next team prepares an event in a similar area, it starts from what was already improved and what held. When someone describes a similar problem, this event comes up among the similar solved cases. Every method writes into the same base, and searching it is what the Brain page is about.
Data stays in the EU, never used to train AI models: Security & Architecture.
Why plants lose what they learn: Europe's productivity is stuck in spreadsheets.
Questions about kaizen
Is kaizen the same as continuous improvement?
Not quite. Continuous improvement is an ongoing process across the whole company, known in German plants as KVP, and a kaizen event is one event inside it: a team improves one area over a few days. A suggestion scheme is a separate system, in which one person hands in an idea.
How does a kaizen event run?
In five steps: prepare, observe, analyse, improve and standardise. The team works in one area for a few days, and the preparation with purpose, target and baseline starts a few weeks before. The event ends with a new standard that has an owner and a review date.
How do you keep kaizen gains from fading after the event?
Write the new way of working down as a standard straight away, and give it an owner and a review date. The review shows whether the gain held. Every action still open after the event has its own owner, so nothing is left hanging.
When do you need a kaizen event, and when is Just-Do-It, PDCA or DMAIC the better fit?
Pick kaizen when one area has several kinds of waste and the team needs to see them on the spot. Just-Do-It is for the obvious fix at one workstation, PDCA for testing one idea, and DMAIC for a deviation that keeps coming back and whose cause you will not find without data. All four methods write into the same knowledge base.
What does a kaizen event leave for the next team?
The record of the event with its observations, confirmed cause, actions and standard, and the result of the review that shows whether the gain held. When the next team prepares an event in a similar area, it finds the record in the knowledge base and does not start from a blank flip chart.
Other methods
Every team picks the method that fits the problem, and one knowledge base stands behind all of them.
Bring the area of your plant you want to improve to the call.