Just-in-Time Manufacturing
Improve flow without overlooking the cost of disruption.
What just-in-time seeks to accomplish
Just-in-time (JIT) coordinates production and replenishment with actual downstream need, limiting unnecessary work and stock. It depends on reliable processes, quality at the source, and responsive replenishment. Toyota places JIT alongside jidoka in its production system. See Toyota’s explanation.
JIT should not be reduced to an instruction to eliminate all inventory. The relevant question is how much stock, capacity, and supplier flexibility are needed to sustain the required service under realistic variability.
Advantages and limitations
| Potential advantage | When it can disappoint | A practical response |
|---|---|---|
| Less work in process and storage | Supplier or transport interruption stops work. | Set risk-based buffers and examine alternative sources. |
| Faster visibility of defects | An unreliable process repeatedly starves downstream operations. | Improve quality and reliability before shrinking buffers. |
| Shorter lead times | Long changeovers make small batches costly. | Reduce setup time and examine product mix. |
| Lower obsolescence exposure | Demand is seasonal, volatile, or intermittent. | Segment products and plan capacity or inventory appropriately. |
| Clearer pull signals | Poor data or discipline creates missing or duplicated signals. | Define signal ownership, replenishment rules, and audits. |
These are engineering tradeoffs, not guaranteed outcomes. NIST discusses how disruption can expose weaknesses in lean supply chains and why manufacturers should balance efficiency with resilience. Read the NIST discussion.
Try a replenishment estimate
One simplified teaching rule estimates the number of containers as
\[N=\left\lceil\frac{dL(1+s)}{C}\right\rceil,\]
where \(d\) is demand per day, \(L\) is replenishment lead time in days, \(s\) is a fractional allowance, and \(C\) is units per container. Use consistent time units. This deterministic approximation does not establish a service level or optimize safety stock.
The defaults give 10 containers, or 300 units of nominal container capacity. Increasing lead time to four days raises the estimate to 20 containers. This illustrates sensitivity to lead time, not a recommendation to carry a particular inventory.
A senior design investigation
Measure arrivals, processing times, setup times, defects, and replenishment delays. Develop alternatives for layout, container size, release signals, and buffers. Compare baseline and proposed performance using measured trials or a validated simulation.
Include disruption scenarios: a late shipment, equipment failure, demand surge, and quality hold. Report throughput, work in process, waiting time, stockouts, and total cost. Demonstrate whether improvements survive variability rather than only showing the average case.
For stable long-run flow, Little’s Law relates average work in process, throughput rate, and flow time: \(WIP=TH\times CT\). Define system boundaries and matching units before interpreting the relationship.
Continue to engineering economics and testing and assessment.