Discrete-event simulation fundamentals for modeling manufacturing and service systems: when simulation beats analytical methods, the entity/resource/queue building blocks, statistical distributions for arrivals and service times, and validating a model against real system data.
The earlier queuing and Little's Law calculators in this studio solve for average behavior with closed-form equations — genuinely useful, but built on assumptions real systems routinely break: multiple interacting resources, non-exponential timing, conditional routing, and interest in a specific shift's behavior rather than a long-run average. This module builds the discrete-event simulation mental model for exactly those cases — entities, resources, and queues; the event-calendar mechanism that advances simulated time; how to choose a defensible statistical distribution for arrival and service times instead of quietly averaging variability away; and a worked single-server queue traced by hand, event by event.
By the end of this module you should be able to explain why a simulation model can be fully verified — bug-free, doing exactly what its logic specifies — and still be invalid if that logic rests on a wrong assumption, and why multiple replications with a documented warm-up period, not a single run, are what earns a simulation result the right to inform a real decision.