Takt, flow and value-stream mapping

Customers need one order every two minutes, yet a supervisor asks every station to work at maximum speed. Local productivity appears strong while orders accumulate between stations. Lean reasoning separates the pace of demand, the speed of one station, and the time a particular order spends in the system. Takt time equals available production time divided by customer demand in the same period. With 480 net minutes in a shift and demand for 240 units, takt is two minutes per unit. Exclude planned breaks from available time, and define whether demand means units or multi-line orders. Do not compare a unit-minute figure with a daily order count.

A value-stream map shows information flow above and material or order flow below. Each process box records cycle time, uptime, changeover and defect rate when relevant. Between boxes, record work waiting and the duration of its wait. If an order waits 180 minutes before checking and 90 before packing, while actual processing is 1.4 + 2.4 + 1.6 = 5.4 minutes, approximate lead time is 275.4 minutes, assuming no other waits. Direct processing occupies at most 5.4 ÷ 275.4 ≈ 1.96% of that time. Even that is not an automatic value-added classification: a check may be essential to safety or customer requirements without directly transforming what the customer buys. State the value test before classifying steps.

The map reveals different wastes. A queue of 120 orders before checking is work-in-process inventory. Moving boxes between buildings is transportation. Reprinting a wrong label is rework. Repeated inspection may contain an immediate risk but does not fix its source. Unused space is not automatically waste without a concrete alternative use. Going to gemba means observing what really happens and asking workers how it differs from instructions, not making a ceremonial visit to confirm a manager's theory. A spaghetti diagram of movements is useful when distances and crossings suggest avoidable travel.

In this fictional case, the team can first measure where the queue forms before checking, then test divided work or removal of duplicate entry without dropping the quality check. Continuous one-piece flow may fit adjacent stable steps. A pull system limiting work in process may fit when steps cannot be directly joined. Just-in-time does not mean abolishing every safety buffer; the decision depends on demand variability and supply reliability. After the trial, monitor defects and lead time together so that speed is not bought at the cost of accuracy.