1 minute read

Imagine that you are an operator working a woodcutter.

What really is the value-adding labour and what is waste?

Really only the cutting part is the value-adding labour. All the things around it are wastes.

Lean now asks you to think about how we can do more of the value-adding and less of the non value-adding.

So I thought to myself: What are my own value-adding and non value-adding activities as a machine learning engineer?

To find this out, I recorded myself and later went over the recording and wrote down every activity that I saw myself doing.

I then took this list and categorized it based on whether it was a value-adding or non value-adding activity.

I was really shocked.

I realized that really the only value-adding thing really is changing code.

But I also realized that I did a lot more than that.

I realized that I spend a lot of time looking for the right information.

I realized that I spend a lot of time communicating.

I realized that I spend a lot of time just simply scrolling through my files trying to find the right file to change.

All of these are wastes if you think that changing code is really the only value-adding thing.

We have surfaced the problem successfully. The waste is clear as day.

Now the question is: How do we do more of the value-adding thing and less of the non value-adding things.

(Note: I simplified a bit. I believe that mentoring others become good at changing code is also value-adding. But I left that out.)