Just-in-Time

Definition

Just-in-time means the right parts reach the assembly line at the moment they are needed and only in the amount needed, so a fully realized flow approaches zero inventory. Ohno achieves this not by scheduling every process forward from raw material to finished car, but by reversing the direction of information: a later process goes to an earlier process and withdraws only what it needs, when it needs it, and the earlier process makes only enough to replace what was withdrawn.

In the Book

Ohno introduces just-in-time as one of the "two pillars" of the Toyota production system in Chapter 1, alongside autonomation. He describes arriving at the idea by inverting the conventional material flow: instead of parts pushing forward from an early process toward final assembly according to a central plan, he imagined the assembly line pulling parts backward from each preceding process, one step at a time. This reversal is what produced kanban — the signboard that tells an earlier process what was withdrawn and how much to replace — but the book is explicit that kanban is only "the means" or "tool" for realizing just-in-time, not the idea itself. Ohno also uses a baseball analogy (Chapter 2): just-in-time is the practiced teamwork of players executing a coordinated play, while autonomation is the individual skill of each player. He is candid that applying just-in-time to a product built of thousands of parts across countless processes is "extremely difficult," and that conventional management methods do not work for it — the whole system had to be redesigned around synchronized withdrawal rather than forecasted push.

Why It Matters

Just-in-time reframes a coordination problem: instead of every upstream station guessing correct output levels from a forecast and pushing on faith, the coordination signal comes from the one place that actually knows demand — the point of consumption — and propagates backward through the chain. This collapses the need for large buffers between steps, but it also removes the safety margin those buffers used to hide: any variability, defect, or breakdown anywhere in the chain now surfaces immediately instead of being absorbed by inventory. The concept generalizes to any multi-stage system (staffing, software delivery, supply chains) where pulling from real downstream demand, rather than pushing from an upstream forecast, trades buffer-based slack for tighter synchronization and faster exposure of problems.