The gist of this article is that you should hold just enough WIP to meet your production requirements with the changeover times you currently have and protect your bottlenecks against malfunction in other resources. So far, this is stating the obvious, and a visit to a Toyota plant or even dealership is enough to see that the Toyota system is not one with zero inventory. You see shelves of stampings, bins of bolts, and trees of wire harnesses. The Kanban system involves some inventory, and, in fact, the only approach that doesn’t is just-in-sequence. What is considered waste is not all inventory, but unnecessary inventory, accumulated for no valid reason anyone can explain.

The article, however, goes further and asserts that it is cheaper to accumulate WIP than to expose and solve the problems that make it necessary, which is a return to the mass-production thinking that was prevalent in pre-Lean operations management.

What the Lean successes of the past decades have shown is (1) that the overall costs of WIP were understated and (2) that the ingenuity of production people and engineer was underestimated. You operate today and next week with the resources that you have, dysfunctional as they may be, and you hold WIP as needed to sustain production. As you do this, however, as an organization, you keep working at solving your problems so that you need less and less WIP month by month and quarter by quarter. This perspective is missing from the article.