Throughput

Throughput is the number of units a production system completes per unit time (hour, shift, day); it expresses the system's actual production performance. Throughput differs from capacity: capacity is the theoretical ceiling, throughput is the actual output. There are two fundamental mathematical ties: (1) Little's Law: WIP = Throughput × Lead Time (Little 1961) — at constant lead time, throughput governs WIP; (2) the bottleneck principle (Goldratt 1984): throughput = the capacity of the bottleneck station, i.e., throughput = 1/max(cycle time). Together these two laws make system performance readable from the triangle of WIP, cycle time, and the bottleneck. The Theory of Constraints (TOC) offers Goldratt's 'Five Focusing Steps': (i) identify the bottleneck, (ii) exploit it (use it to the maximum), (iii) subordinate everything else to its pace, (iv) elevate the bottleneck, (v) move on to the new bottleneck. Ways to increase throughput: reduce bottleneck cycle time, add parallel stations, add shifts, shrink transfer batch (transfer batch ≠ process batch distinction). In SMB practice throughput comes from daily production reports, MES data, or simple per-station output counts. Annual throughput × unit contribution margin is the basis of throughput accounting. Job-shop scheduling, flow-shop scheduling, assembly line balancing, and aggregate production planning all sit on its output side. References: Goldratt and Cox 'The Goal' (1984); Goldratt 'The Theory of Constraints' (1990); Hopp and Spearman 'Factory Physics' (1996); Little (1961).