Same with lot sizes as you mentioned. I have an online calculator for buffer sizes and wanted to write an article about lot sizes for quite some time now. But no promises on when I get around to write it 🙂

To this day, concepts like one-piece-flow, lot size 1, buffer inventory, and batching have way too much dogma around them to the point where it overrides common sense in case of obvious conflicts.

This gets tricky when you have a mix of processes that are discrete from each other, work on batch inputs e.g. a dishwasher, and processes that are complicated to stop and restart.

Hands down, the absolute #1 benefit of one-piece flow is the reduction of half-finished parts stacked around i.e. WIP and all of the associated wastes that go with it e.g. things have to be moved around when there’s a switch.

Another case would be a subassembly line made up of however many processes that produces subassemblies that are used by the main line. It’s a completely sensible thing to do to have the output subassemblies to be placed onto a pallet until it’s full and a forklift grabs it for delivery to the main line. Sure, it would be awesome to relocate the entire line so that the finished sub-assys are fed straight to the receiving line without extra transportation. Whether that’s even practical or worth the time and expense is a different story.

Batching is another concept that often conflicts with the dogma. It’s a silly example, but how about cooking bulk foods such as rice at home. Do the purists insist on preparing and eating rice one grain at a time? With enough money and inclination, it’s entirely possible to build a machine that will prepare rice on a grain-by-grain basis. But should we? Or is setting the batch size to a reasonable amount e.g. prepare a dinner’s worth at a time vs preparing the entire 50lbs bag more than good enough?

It would be cool to see a post on how to choose lot sizes and the nuances of where buffer inventory becomes overproduction, and vice versa. It’s very common urge to reduce extra inventory to nothing in the beginning, and some overproduction gets traded in favor of daily firefighting.

I think the the real benefit is that moving to one-piece flow exposes all the problems in the system that make it hard to make/move one piece at a time. I think that is the beauty of it. It requires excellent quality and no waste.