Do You Really Need to Save Seconds in Scheduling? YES!

Want to know how to translate a second into dollars? It all comes down to calculating your scheduling down to the second.

Many applications don’t even use minutes to compute schedules. They use time “buckets” of five- or 15-minute intervals–or even hours or days. You’ll see this “buckets of time” approach most commonly in spreadsheet-type scheduling applications because each column in a spreadsheet can be used to represent an interval of time. But, as I’ve written in a previous blog, (Click here), time buckets provide computational conveniences for the application developer, but they fail to produce good schedules for the end user–certainly not schedules that approach maximum capacity. They might be suitable for appointment booking scenarios–like in a doctor’s office–but they have severe limitations in manufacturing, fabrication, and project scheduling environments.

Let’s say that the ACME Manufacturing Company makes widgets using several different production sequences. But orders for widgets arrive in all different quantities. The manufacturing sequences to produce any order are similar, but big orders obviously take longer to produce than small ones.  Wouldn’t it be helpful if the scheduling application allowed you to define the sequence for a minimum order and then stretch the sequence as necessary for larger ones? You certainly wouldn’t want to define different production sequences for every different order size. And you might not want to restrict your manufacturing procedures by defining fixed-lot sizes. That approach leads to inventories that are too big and cost too much to maintain.  You don’t want to over- or under-produce because your scheduling application restricts you to fixed-sized lots.

A minimum-sized production sequence for producing a single widget might need to have activities defined in terms of seconds, even though you would never make one widget at a time. Then a sophisticated scheduling application could use the minimum-sized sequence as a template for realistically-sized order quantities. Assuming run times are directly proportional to quantities to be produced, the creation of a new production run is simply to select the proper production sequence and specify the quantity. The activity durations and resource requirements can be scaled to the specified quantity. Times specified in seconds within the template become accurate durations in minutes and hours for the real production run!

Now here’s a detail that’s important: Some activities in the template sequence may have durations that are not dependent on quantity. For example, a set-up task for a work center may be done in the same amount of time whether the quantity to be produced is a few or many widgets. The same may be true for a change-over or clean-up task. These tasks should not be scaled up by the quantity to be produced. So, a sophisticated scheduling application should allow you to specify which activities in the template sequence are to be scaled and which ones are not. The difference is important for correct modelling of the production capacity.

Modelling in seconds and using proper scaling logic translates directly into dollars saved—or earned. Does your scheduling app do that?