Scheduling may be the oldest profession, at least in a food plant, but it is one of the most difficult to understand and manage. It’s a little like the weather, everyone complains about it, but it’s hard to make it better.
The plant scheduling function is at the intersection of planning and execution. If the planning process - forecasting, demand planning, sales order processing, and inventory planning - fails, the plant schedule has to be changed. If the plant execution breaks down - shortage of ingredients, equipment failure, lines running slower than plan, or labor shortage - the plant schedule has to change.
We were discussing scheduling improvement strategies recently with Tom Crowell, project manager information services at Dreyer’s Grand Ice Cream. Tom reminded us that, while both the planning and execution processes are certain to occasionally fail, the key is to have a scheduling process that can quickly react to changing realities.
Today, when we think of improving reaction time, we look to automating the process. Selecting the right scheduling tool is important, but we must also look at the scheduling process itself. If your plant scheduling is still managed on spreadsheets or if you find that the schedule is in perpetual motion, you might consider developing a scheduling improvement process.
If you find that most schedule changes are generated by the planning process, then your planning assumptions - inventory levels, shipping times, throughput rates - should be re-evaluated. If plant performance is the key factor, then the relationship between what is being scheduled and what actually can be made must be adjusted.
According to Gil Garcia, vice president of business development at SCT, “The quality of the data coming from the planning and the execution system is critical to the accuracy of schedules.”
One factor that makes the scheduling process hard to evaluate is that it can be set up at increasing levels of detail. The most general production schedule would be a plant loading - over a fixed period of time the plant is scheduled to produce X amount of product. More detail might be a line loading where products are sequenced on lines over a fixed period. At the lowest level of detail, components within a line could be scheduled separately, and then combined to make a line schedule. The level of detail in the scheduling process should be matched to the level of detail in the production environment.
“The model of production used for scheduling must match the realities of production,” according to Garcia. “For example, changes in process equipment or seasonal changes in raw materials can have a significant impact on line rates. When the projected run rates on the schedule are not regularly adjusted to match the actual capability of the production lines, the scheduling process will be problematic. Many of our food customers even sequence schedules based on quality characteristics, such as making the garlic bagels last, so they don’t contaminate the plain bagels.”
There is not a shrink-wrapped method of improving your scheduling process. Lots of the frustration generated by plant scheduling comes from a mismatch between the level of detail plant operations requires to run smoothly and the level of detail available in the scheduling system. When companies take the time to understand their scheduling requirements, they can develop organizations and systems that will make scheduling an integrated part of their business. But it is essential that operational staff trust the schedule. The key to success here is good data.