Why the Fastest Step Does Not Set the Pace

Imagine three separate stations handling identical fictional paper jobs. Every job must visit station A, then B, then C. A needs two minutes, B four and C three. Each station handles only one job at a time, but different stations may work concurrently on different jobs. Jobs wait in order without overtaking.

All six jobs in the later batch are ready at time zero. Processing starts immediately when both the job and its next station are available. For this exercise there are no transfer delays, breaks, setup, rework, faults or competing tasks. Those are deliberate exclusions, not assertions about ordinary workplaces. No person is asked to match these invented times.

Draw three boxes and arrows in the required order. Put the time and its unit under each box. A number such as four is meaningless until you say whether it is minutes per job, jobs per minute or the total number of jobs. Write the concurrency assumption beside the diagram because it changes what adding the times can explain.

For one job with all stations initially free, processing takes two plus four plus three: nine minutes. That is an individual job's elapsed processing time under these conditions. It does not yet describe the interval between repeated completed jobs or how long six jobs take. Defining the route, units and availability first prevents a tidy formula from answering the wrong question. In actual work, separately identify permissions, safety checks and excluded delays before interpreting any timing model.