Just a note that the article (perhaps inadvertently) confuses a little: industrial manufacturing <i>must</i> be highly routinized and boring if it's working well. Tolerances are very tight for well made manufactured goods, since they have to be to be assembled into larger, more complex goods. Let an operator get creative with a running process, and all bets are off [0].<p>Besides! People are also <i>terrible</i> at that sort of work, so it's better to automate the job and have the people supervise machines. The machines do the terrible soul-crushing tedious work and the worker oversees the hiccups and glitches. (Machines are pretty lousy supervisors, and are difficult to program supervisory roles in the messy situations a lot of manufacturing happens in.)<p>And keep in mind that's for a <i>running</i> line or process. I don't think there's room for creativity on carrying out the line's program, but there's endless opportunities for planning and improving the process off-line. People are great at deliberation, planning, and brainstorming. (<i>Too</i> good, some may cynically say :) I.e., you'll never get a better ROI than buying a group of operators lunch and let them rail at you about something that irritates them about a line. And include them in planning/upgrade meetings; they own the line, after all!<p>And there's the exception of those times where it just makes economic sense to plop a person down for something boring and lonely. For example, the cost of vision control tech is getting ever more affordable and powerful, but it still takes a substantial amount of engineering work to actually install and tune the system for the specific situation it's in; by contrast, people are cheaper in the short-run and can be reasonably effective very quickly. (It's mind boggling when you see a failure that you think, "I could have paid a guy a year's salary to sit in a chair and watch for that one obvious, uncommon failure mode event.")<p>[0] This is an interesting dichotomy: the workers <i>must</i> own the process they're working on, but they also need to not constantly fiddle with settings. If they're not <i>actively</i> tweaking stuff, then they feel disenfranchised, and if they are constantly messing with stuff then the operator is simulating a randomly tuned PID loop on whatever dials happen to be nearby. The root is that people are generally terrible at figuring out the difference between "normal variation" and "exceptional variation", and tune both. So when Deming talks of <i>knowledge of variation</i>, this has to be balanced by <i>knowledge of psychology</i>.<p>And that balancing act brings us to this article's contention.<p><a href="https://en.wikipedia.org/wiki/W._Edwards_Deming#Key_principles" rel="nofollow">https://en.wikipedia.org/wiki/W._Edwards_Deming#Key_principl...</a>