Every moulding shop I have worked in had a lessons learned file.

Some kept it in a binder, some in a spreadsheet, some in the closing report of every program. Ask when anyone last opened it before a design review and the room usually goes quiet. Lessons learned in injection moulding.

That is not because people do not care. It is because the file sits in one place and the work happens somewhere else. The lesson gets written at the end of a program, when everyone is tired and already moving to the next job. Then it gets filed. Then the next mould gets built with the same horn pins, the same insert design and the same problem.

The binder nobody opens

For years we tracked this kind of thing in meeting minutes and closing reports and considered it done. We logged who said what, when and why. It turns out nobody reads meeting minutes in the rush of a busy day, and nobody reads them at all during the launch of a new program.

So the lessons existed. They were written down. They just never reached the person who needed them at the moment they needed them.

That is the pattern I keep coming back to. It is rarely a knowledge problem, it is usually a time problem. The knowledge is in your building. What is missing is a way to capture it and use it that costs less effort than the people on your floor have to spare.

Where the real lessons come from

Most of the best lessons I ever learned did not come out of a lessons learned meeting. They came from the shop floor.

Ron, a setup guy with 30 years on tools, watched his crew spend six hours pulling a mould apart just to change inserts while the press sat empty for a full shift. “No tool should have to be taken apart to remove an insert,” he told me. “Makes no sense. Always front load.”

Dave, our maintenance guy, pointed me at a bin on his workbench spilling over with broken horn pins and asked why we were not using locks, and why the pins loaded from the back.

Mike, a toolmaker at a different shop, told me to use Unbrako bolts on a badly mounted ejection cylinder. Then he added the part that showed his real experience: do not go too hard, because if one fails he needs to be able to drill it and get an easy-out in.

Each of those lessons was worth real money. Each one came over the noise of a press, in a sentence or two. And I could barely keep up with writing them down, and I had the rest of my job to look after. If a lesson depends on an engineer finding an hour at a desk later, a lot of them never make it.

Make capture take seconds

The first fix is to make capturing a lesson almost effortless. If it takes 2hr to write a proper entry, it will not happen on a program that is running behind, which is most of them.

What works is catching the lesson where it happens, in the words of the person who has it. A voice note on a phone that drops into a shared database. A photo of the broken part or the fix. I used photos of Mike’s work to set our best practices, and a photo carried more than a paragraph would have.

A useful lesson only needs four things:

• What happened, in one or two sentences.
• What we did about it.
• What to do differently next time.
• Which task or stage it belongs to: part design, mould design review, tool trial, setup or preventive maintenance.

The last line is the one most files skip, and it is the one that makes the next step possible.

Put the lesson in front of the right task

A database you have to go searching through is just a binder with a search box. People are busy. They will not go looking for a lesson they do not know exists.

The lesson has to come to them. Not sitting in the moulding standards or a LL database. When the mould designer is specifying horn pins, Dave’s lesson should be sitting right there: front load, with locks. When the insert layout is being reviewed, Ron’s lesson should be on the table before anyone signs off. When someone is choosing fasteners for an ejection cylinder mount, Mike’s caveat about hardness should show up with it.

That is why the stage tag matters. And it is why the design stage is where lessons pay the most. A lesson that surfaces at the mould design review costs a line on a drawing. The same lesson discovered at the press costs a pulled tool, a lost shift and everything else on the schedule pushed back.

I have been building toward this for a while. When I wrote about AI in project management, I said the linking part, the tools that put the right lesson in front of the right task, was not reliable yet. Now my checklists are updated by emails and the checklist automatically front load into the check lists and the checklists are automatically issued to the designer/tool maker/manufacturing. Oh how things have changed since the 2hr for a registering of a lessons learned and lost information.

Start manual, then automate

You do not need software to start. Here is a version any shop can run manual next week. Remember every successful complicated system was born from a successful simple system. Build test prove manual then automate.

Keep one shared file. Every entry gets the four lines above.

• What happened,What we did about it, What to do differently, Which task or stage it belongs to

At every design review, mould design review and tool trial, take ten minutes before the meeting starts and read out the lessons tagged to that stage. Ask one question: does any of this apply to the job in front of us?

When a lesson proves itself on two programs, move it onto your design checklist so it stops depending on anyone remembering. That is how a checklist grows from real experience instead of from a template. Also keeps the checklist from becoming meaningless.

Once that habit is working and the file is too big to read through by hand, that is the time to look at tools that do the matching for you. Every working automated system started as a working manual one.

Show people their lesson got used

There is one more piece, and it is the one that keeps the whole thing alive.

When we put Dave’s front-load horn pins with locks into our tools, the whole maintenance department took credit for the improvement. They started offering more suggestions because they could see they were being heard. When Mike saw his work in our best practices, he started bringing me more ideas than I could keep up with.

People keep sharing lessons when they see them used. They stop the day they feel their advice disappeared into a file. So when a lesson changes a design, go back and tell the person who gave it to you. I remember the skin score operator describing to me his operations. He was very proud of his efficiencies and methods that he used in his job. Some of the things were unique to that operator and freedom is needed for individualism just meet cycle…. but one thing he told me opened up a door to a beautiful solutiion in a very difficult time developing new equipment with 1/4 of the time required. I went back and let him know when the new equipment arrived that meeting the customers compressed schedule was largely due to the discussion that he and I had.

Where to start

Pick your next design review. Before it starts, ask the setup lead, the maintenance lead and the toolmaker for the one thing they wish the last mould had done differently. Write each answer down with the four lines.

• What happened,What we did about it, What to do differently, Which task or stage it belongs to

Read them out at the review.

That is a lessons learned system. Everything else is making it easier.

I have spent 35 years in moulding and product development and hold a PMP. Most of what I do now is help moulders front-load the design stage and keep hold of the knowledge already in their shop, as a project manager when you need one. If your lessons learned file is mostly a binder nobody opens, I would be glad to talk it through.

Get in touch:https://launchpadprojectmanagement.com/contact-us/