Analysis of the effect of learning on the throughput-time in simple assembly lines

Calculating the throughput-time of a production run, when a learning effect influences task times, is a complicated but important problem of production management. Classical assembly line balancing techniques assume a constant cycle time. However, if a learning effect is present, cycle time changes for two main reasons. First, cycle time continuously decreases as the operation time of the bottleneck station decreases due to learning. Second, the bottleneck may shift from one station to another, causing further changes in the cycle time. This paper explores the effects of an exponential learning function on the operation of simple assembly lines. An algorithm is presented to determine throughput-time, and a sample problem is provided to illustrate the required calculations. The sensitivity of the throughput-time with respect to the learning rate is also examined, and some general conclusions related to the effect of learning rate changes are presented.