A method for reducing inter-departure time variability in serial production lines

The inter-departure time variability is an important measure in production lines. Higher variability means added work-in-process and less predictability in output. It can be a primary obstacle towards achieving on-time delivery. The effects of line parameters (e.g., line length or buffer capacity) on inter-departure time variability have been studied in recent years but no method has been proposed for its reduction. In this paper, such a strategy is proposed and studied via simulation. Results indicate that significant reductions (of more than 20%) in inter-departure time variability can be achieved for as little as 0.5% increase in the mean inter-departure time or without any increase at all, for a majority of the line parameter values experimented. This was found to be the case for symmetrical (uniform) processing time distributions as well as for asymmetrical skewed (exponential) distributions. Similar results have also been obtained in the application of the proposed strategy for the case when one station has a higher variance than the others. Therefore, in situations where output predictability is more of a problem than capacity, this strategy constitutes an effective alternative.