When a launch goes sideways, the post-mortem usually lands on the same conclusion. We did not know. Nobody saw it coming. We need more training, better engineers, a smarter system.

In my experience, that is almost never true. Most of the time, somebody in the building knew. The problem was what it cost to get that knowledge from where it lived to where the decision was being made.

I have two stories that make the point better than I can argue it.

The knowledge was there both times

The first is the defroster grille program I have written about before. Our FEA engineer, with fifteen years behind him, told us the mounting points would crack, the parts would warp, and the mould flow showed choked flow and uneven cooling. He listed exactly what to change. We had the knowledge on paper, before we cut steel. We built the mould anyway, ran 25 percent scrap for three months, and paid $800,000 for a replacement that included every change he had asked for.

The second is a better story. I was stuck trying to stack several moulds on one platen for a pool system. The external water lines took up too much room, and I had resigned myself to running the job on two presses. Dave, our maintenance guy, said seven words over the noise of the floor: “Ya know, if ya did internal manifold it would fit.” That suggestion saved the customer $2 million over the life of the program, and every mould we built after that used internal manifolds.

Same pattern, different ending. In both cases the answer already existed inside the company. Nobody needed to go learn anything new.

What a time problem looks like on a moulding program

When I say time, I do not mean the hours on a time card. I mean the time it takes to get a piece of knowledge out of one person’s head, or one report, and into a decision. Every step costs time, and some cost money, attention or nerve as well.

On the grille program, using the knowledge meant a $50,000 quote and eight more weeks on a schedule that was already tight. That was the time cost, and on that day it felt bigger than the risk. So the knowledge got acknowledged and set aside.

With Dave, the cost was different. That shop had aggressive leadership, and most people had learned that speaking up was not worth it. For Dave, the cost of offering an idea was the risk of being shot down again. What lowered it was months of me working the floor in my own beat-up boots, doing rounds with mould repair and working shifts with the millwrights. By the time he spoke, it cost him almost nothing to say it.

That is the difference between the two stories. Not what people knew, but what it cost to use it.

The time it takes to capture it

Here is the part I am less proud of. After Dave, there was Ron, a setup guy with 30 years on tools, who told me no mould should come apart for six hours just to change an insert. “Always front load.” There was the two-foot bin of broken horn pins Dave pointed at, which led us to front-load horn pins with locks. There was Mike, a toolmaker who taught me to use Unbrako bolts on badly mounted ejection cylinders, but not ones so hard he could not drill them out if they failed.

Every one of those was gold. And I could barely keep up with the documentation, because I had the rest of my job to look after.

That is the time problem in its plainest form. The knowledge was flowing. The cost of writing it down, organizing it and putting it in front of the next person who needed it was more time than one busy engineer had to spare. Knowledge that is not captured walks out the door when people retire or move on.

A good example is a safety feature that never made it into our documentation. The safety department required a simple gauge on the mould, but it wasn’t written into our best practices. The mould got built. The toolmaker wasn’t aware of the requirement, and even if they had been, the tool designer wasn’t. The tool wasn’t checked against a standard checklist, and it arrived in-house missing a critical safety feature.

That is where I decided enough was enough. I had to get better. We set up checklists, created a lessons learned record, and front-loaded those lessons into the checklists. Ultimately, that is how the LPM Suite was born.

Make the quick path the right path

If the problem is time, the fix is not more effort. People are already busy. The fix is making the right thing quicker than the wrong thing. A few ways that works in a moulding shop:

– Capture in seconds, not in reports. A voice note on a phone that lands in a lessons database will get done. A form back at the office will not. It takes me about 2minutes to capture the lesson in my database through voice recorder on my phone. I later collect the voice files, convert them and then OpenClaw does the rest. Maybe 10minutes to batch it for now.

– Bring the lesson to the task. The front-load insert lesson is worth nothing in a binder. It is worth a lot when it shows up at the start of the next mould design review.

– Price the warning at the moment it is raised. Put the cost of acting next to the cost of being wrong. $50,000 and eight weeks looks different beside $800,000. I know this is easier said than done but when you explain to the toolmaker what you are doing the tool maker can usually get a pretty good ball park quote within 10% fairly quickly.

– Lower the cost of speaking up. Be on the floor often enough that the setup guy does not have to work up the nerve to tell you something.

None of this is complicated. What makesshop floor knowledge capture hard is that each step has to be easier than skipping it, every time, on the busiest week of the year.

Where to start this week

Pick one recent problem in your shop that got solved on the floor. Ask the person who solved it to tell you about it, and record it on your phone while they talk. That is one lesson captured for the cost of a five-minute conversation. Free AI can process this. 

Then ask where that lesson needs to show up next time. A design review checklist. A mould build spec. A note to the toolmaker. Put it there. This is LPM Suite, AI can process this. The AI tool is how doing the right thing gets faster than doing the wrong thing. That is what has changed.

The design review checklist is the most critical place for items to be caught; catch it in design 10times cheaper. After that the mould shipping checklist is the next best spot. Through all the chaos and change in a mould tool / part design evolution, the mould shipping checklist is your last chance to get it right. Nothing worse than having to ship tools back to the toolmaker simply due to an oversight.

Do that a few times and you will see how much your shop already knows. Most moulders I talk to do not need more knowledge. They need less friction between the knowledge they have and the decisions they make.

shop floor knowledge capture

That is the work I do. I have spent 35 years in injection moulding and product development and hold a PMP, and I help moulders front-load the design stage and put the knowledge already in their shop to work, as a project manager when you need one. If this sounds like your shop, I offer a free Diagnostic Map conversation to find where knowledge is getting stuck.

Let’s talk