Factory Flow: Capacity, Variability, Rework and Bottleneck Migration
How the Constraint & Flow Lab turns local rates into a system-level flow argument.
What the showcase models
The lab is a deliberately small serial production system. Each stage has finite effective capacity; demand, variability and rework alter the load placed on the system. The governing constraint is the stage with the least effective headroom under the scenario.
Flow relationships
At steady state, throughput cannot sustainably exceed the limiting effective capacity. Little’s Law, WIP = throughput × cycle time, provides the conservation relationship connecting inventory and time. Rework increases effective workload because some units consume capacity more than once.
Controls and interpretation
The point of the controls is not to maximize every station. It is to watch the constraint migrate. Adding capacity to a nonconstraint may create little system throughput; relieving the active constraint can expose the next limiting stage.
Assumptions and limits
The browser model is a deterministic teaching model rather than a full discrete-event factory simulation. It compresses distributions, blocking, starvation, calendars, product mix and detailed routing into simplified effective rates. Use the DES lab when event timing and stochastic queues are decision-relevant.
Validation checks
Increasing capacity at the unique bottleneck should not reduce theoretical system capacity. Increasing rework should not improve effective capacity. If demand remains below every stage capacity, the model should not invent a capacity-constrained throughput loss.