The introduction rate of new SKUs into the food chain continues to grow at an unprecedented rate. Increased numbers of SKUs means increased complexity with scheduling.  Increased complexity with scheduling means ever more complicated spreadsheets that may be obsolete shortly after they are published.

If a push system like MRP or MPS is in place there may be a constant flurry of cancellation, expedite, or delay messages/actions to be dealt with. Schedule thrash is induced for a number of reasons including, line personnel no shows and skill gaps, customer order changes, late or incorrect supply, forecast revisions, unplanned downtime, unplanned rework, orders triggered by reorder points on safety stock, and much more.

In short, it is complicated. 

Advanced planning and scheduling systems are not necessarily a panacea either.  For these systems to work we have to define the business rules around the schedule. What can be scheduled after what and which machines for which recipe. The more diversity you have within your recipes and the more dynamic they are the more of a challenge to utilize the tool.

To deal with this complicated problem we have to attack it from multiple angles. Uncertainty of demand may be buffered either with production capacity or inventory.  A business decision must be made for each customer on what response level we are willing to invest in. The closer we get to 100% success on meeting demand the more the buffering capacity or inventory increases. To improve schedule stability we have to go at the problem from both ends. 

First, need better partnerships with our customers to avoid last minute orders and abnormally small order quantities. We need to ascertain the costs of the schedule volatility and use that as input into business relationship and pricing decisions. In short, we need to improve the supply chain collaboration. 

Second, we can improve schedule volatility by identifying those SKUs with good history that are reasonably forecastable. The forecast can be used to set reorder points using Gordon Graham reorder point calculations. That calculation is based on Reorder Point = Manufacturer’s Average Lead Time X Average Daily Usage X Safety Stock Multipler.  We advocate using a safety stock multiplier that factors in the historic variability of demand.  Properly tuned the system will trigger production/purchase orders to replenish safety stock and relieve some of the scheduling volatility.

Third, we need to segment those products that are not very forecastable or whose shelf life is too short to effectively stock product.  We need to look at the aggregate demand for these products and make a decision on the amount of capacity we are going to keep on hand. The schedule can be organized into slots based on product types. In these situations we need to generate raw material forecasts based on the anticipated product type demand.

In summary, schedule improvement requires a partnership between sales and operations.  Each team needs to understand the given constraints. Investments need to be balanced with customer commitments. Adroit has the background and experience to work with your team to systematically attack this problem and drive improvement.