The Prepared is an excellent newsletter. In this latest issue Matthew Hockenberry wrote a brief but thought provoking summary of the development of just-time-time manufacturing.

This is just-in-time production: “making only what is needed, only when it is needed, and only in the amount that is needed.” The result, Toyota claimed, was not only a radical reduction in waste, but in costs associated with financing and storing inventory. It should be obvious, of course, that this system required some underlying structural changes. The first was linking the pace of work to the pace of sales, and a reliable means of distribution between the factory and the retailer. The second was a network of suppliers and partners who were also operating “just-in-time.” While the first requirement is explicit, the second is not. Indeed, Toyota goes to great lengths to argue that it isn’t a requirement at all. But just-in-time is contagious, and when a company adopts it, supply contracts become smaller and shorter, demanding a responsiveness that is difficult to achieve without suppliers operating in the same way. After all, just-in-time production affords the flexibility to make frequent changes throughout production cycles, or to correct “defective” runs – something that is challenging with large, static inventories. The risk of holding parts that may no longer be needed by your largest customers means most suppliers soon came to adopt at least some elements of the system.

Beyond a history lesson, there are insights into current supply chain crisis: neither mass production nor just-in-time manufacturing are enough to see us through our current troubles. There is still room for more dynamic (and reliable) forms of production.

Equally important is how just-in-time was adapted to the United States. For many firms, the adoption of lean production techniques in the 1980s and 90s corresponded with increasing computerization of their production process – and the adoption of Material Resource Planning systems in particular. But when software that had primarily been intended to plan was increasingly used to predict, the ideas that had informed Toyota’s work—a responsiveness to the material facts of Kanban cards embedded in the actual machinery of production—became increasingly abstract. They were no longer just just-in-time. They were the imagination of what might-just-be. And when these predictions have failed, and there is nothing to pull from in reserve, the effects have proven catastrophic. When faced with unpredictable circumstances, the industry response has been to shore up models, double down on a future of artificially-intelligent planning systems. Of course, this all assumes a future that we could actually plan for.

Check out the whole post here.