Machine setup times are reduced by streaming jobs together that have a high degree of "matching"; ie. schedule jobs based on the parts they require, and how those part requirements are similar to other jobs. Jobs that have similar parts requirements require less setup between jobs.

You have to look at the average number of jobs in the queue, the number of production lines, the different product configurations, etc., before you can design an appropriate algorithm.

>Hello All,
>I have a Jobs.Dbf (one row per job) with a one-to-many relationship into JobParts.Dbf
>
>The production process is semi-automated with machines that can receive/take parts simultaneously from several sources (feeds), for example: conveyor belts, mini gantries, collection bins, etc.
>
>Operators are spending too much time managing the manual part of the process: supplying parts to the various feeds. Each time a change of part is required the whole process has to stop. Ideally I would like to queue the jobs so that transitions in the Part Feeds are minimized.
>
>Any ideas appreciated
>Pete.