Just finished Taiichi Ohno's The Toyota Production System. It's great. The only problem I have with it is that it's so dense, I'm not likely to get everything from one reading alone.
If you work in software, and have dabbled in agile or kanban or lean, read this.
Conversation
A short thread of ideas from TPS that in no way does Ohno's writing proper justice:
1) The goal of TPS is to reduce waste in any form. And when Ohno says ‘any’, he MEANS any: inventory that piles up is considered wasteful. Excessive speed is considered wasteful.
1
2) Overproduction is waste. So have the latter step in production take parts from an earlier step. Have that earlier step sit idle until it needs to replace the buffer. Don't produce ahead of need. Don't keep inventory. Don't build warehouses.
This is called kanban.
Replying to
3) Efficiency MUST lead to cost savings. If you get a new tool that saves 90% of a worker's time, this is useless because you are still paying for that one worker. Only buy the tool if you can remove one whole worker.
1
4) Automation prevents you from adapting to reduced demand. So don’t blindly automate.
Toyota does something called 'autonomation' — a thoughtful split of tasks between man and machine. That way, they can reconfigure to have less workers when market demand dips.
1
3
5) Depreciation is an accounting fiction. If an old machine that is well maintained is able to produce what is demanded of it … keep it around! It is wasteful to retire it. Tools should be evaluated based on what how much future value they may create!
1
1
There are probably more nuances, and each of the five ideas above interact with each other in subtle ways. It's too much to wrap my head around, and I spent a whole month reading the book!
Totally recommended.
1