One of my favorite customers makes appliances and components for the consumer products industry.
Today we went live with a new concept linking forecast and customer orders with a production scheduling strategy. The key phrase is "smoothing irregular demand".
Our customer's production plant receives orders from each of the networks DC's as well as direct customer orders. The DCs, of course, take the customer orders and that demand can be quite irregular. Up until now the supply chain acted quite random and planners, schedulers and sales reps had a hard time at work.
This, as I believe, was due to the "order related and oriented" process dominating fulfillment, distribution and production. A sales order from the DC was forwarded to the production plant in the same quantity and then translated individually throughout the BoM. A yellow slip followed the order from the raw material through final assembly to packaging, distribution to the DC and finally the delivery to the customer.
In other words: the sales rep entered an order and the supply chain sprung into action - or should I say: "slipped into desparation and chaos"?
So what do you do? First, we asked the planners in the DC to maintain a forecast (strategy 40 for a VSF that gets consumed by sales orders). All these forecasts caused - through stock transport orders - a cumulated demand in the production facility. After that we let the system generate a repetitive order which covers the total demand and then we split the order through a heijunka strategy into EPEI (every product every day). The resulting schedule was leveled onto the final assembly line, carefully ensuring a maximum 75% utilization. Then the direct customer orders had room to be scheduled directly on the line (strategy 81 ensures the direct creation and scheduling of a repetitive order right out of the sales order).
We chose to do this on the final assembly line because it has ample capacity and very short cycle times. It can also be reset in a haertbeat. But most important it ensures the highest degree of customization.
The next challenge was component availability for final assembly. Here we choose SAPs eKanban. With it you have a full container for X items (regular and high demand) and Event-driven or one-time Kanban for Z items. Using this strategy you can be assured to pull high sellers quickly from the line and quote the customer a true lead time for slow movers.
Back to the DC. Since a forecast VSF was maintained, the DC will now receive regular shipments - on time! - to maintain an inventory. This inventory should be available at the beginning of the period and will be consumed by customer orders throughout the period. Therefore forecast accuracy and inventory responsibility remains in the DC and should be controlled with safety stocks (I recommend a coverage profile for dynamic SS).
The resulting demand in the production facility arrives in a regular fashion and gets further smoothed by Heijunka.
Of course there is a lot of setup required - correct lead times, effective lot sizing, right strategy and the appropriate MRP type and replenishment policy - but absolutely nothing that can't be done with standard SAP. Not one Z setting was used along the way..,
Keep it up this post always updated with valuable information regarding on chain assemblies.
ReplyDelete