A production route can only sustain output as fast as its constrained resources allow.

If station $i$ requires cycle time $t_i$, its ideal capacity is approximately

$$C_i=\frac{1}{t_i}$$

in consistent units. For stations in series that each process every unit, the smallest effective capacity limits the line's long-run throughput. That resource is a bottleneck.

Suppose three stations require 1.0, 1.5 and 0.8 min/unit. Their ideal capacities are 60, 40 and 75 units/hour. The second station therefore limits ideal line throughput to about 40 units/hour.

Utilization is the fraction of available resource capacity demanded by work. Very high utilization can seem efficient, but when processing times and arrivals vary, queues can grow rapidly near full utilization.

Improving a non-bottleneck from 60 to 70 units/hour does not raise a line limited at 40 units/hour. Improving the bottleneck can.

Capacity analysis therefore asks:

1. Which resources must every product visit?
2. What effective rate can each sustain after downtime and setups?
3. Where does work accumulate?
4. Does the limiting resource shift after an improvement?

A bottleneck is a system property, not permanently a particular machine. Product mix, downtime or process changes can move the constraint.