Scheduling Assembly Lines with SAP
Many of the companies which manufacture their product with assembly lines (packaging lines, final assembly. sub-assembly) and run SAP, do not use SAP’s functionality to schedule the line. Even though in most cases people talk about takt, they still use discrete production orders and often do the scheduling in Excel.
The thought that SAP can’t handle this type of scheduling is still prevalent and couldn’t be farther from the truths.
In your company you might have one of two types of assembly line execution (there are many more, but here I’d like to focus on the ones that assemble customized products). Either you make everything custom to order and your feeder lines produce very specific, customized semi-finished parts or you have common products that you can use as a buffer in front of your final or main assembly line. In the second case you can use de-coupling points and manufacture to stock… then your assembly line pulls from that stock. this requires separate work orders for the feeder lines. Separate from the work order that runs through the main assembly lines.
For case #1, where everything is customized, this type of planning does not make sense as everything is specific and any buffer of customized parts lies around unused until the time comes for it to be finished for that very specific customer order. In that case the best choice is to create one, and only one routing that contains all the operations including the operations for the feeder lines.
Let’s illustrate this point. For a company that makes products with an overseeable amount of options, like cleaning machines, you can work with inventory points in the value chain. Upstream of the final assembly line you may want to keep buffers of motors, wheels and seats from which your customer orders can pull into a final, customized assembly.
This type of manufacturing might look something like this:
Airplane and automobile production is similar in this respect as you have a very deep Bill of Material with a manageable (and therefore stockable) amount of options on each level.
Now imagine you’re in the business of producing highly customized products where every finished product is ordered to a customer’s very specific requirements. Examples of this type include specialty vehicles like ambulances, RV’s or firetrucks, field power transformers (they are very different from one another but have very similar components) or traditional satellite production. In these cases the customization goes deep – meaning that you cannot stock the intermediates. On a transformer the bushing might be slightly different every time you pull it into the assembly line. So there is no point of stocking various bushings when each bushing can only be used for one specific assembly.
Therefore you want to manufacture that parts and sub-assemblies only when they can flow directly into the final assembly line and that process might look like the following sankey diagram.
Note that it is absolutely pointless to start production of a sub-assembly beforehand. What I have seen is that schedulers think it to be of advantage if they start a portion of the job before the customer has finalized their specification for the final product. What will happen is that the part is finished and will have to wait until it can be used for the final assembly. There is no way it can be used for anything else but that specific job. So why don’t we wait until we have the final order and then start the job and let everything flow together? You might argue that the lead time is longer then. No, it’s shorter becuase mothing has to wait and flows.
What you’ll have to watch out for is that you define a frozen zone that is longer than the lead time to produce your product from the beginning to the end.
If you’d like to run the assembly lines by takt – and that is the only way you create flow then the best choice in SAP is to use Repetitive Manufacturing with its sequencing board and line balancing. However, you can also create ‘takt’ using a discrete order where every operation has the exact same lead time so that you move the product from one work station to the next at exactly the same time disbursement along the line (what you lose with discrete orders is the ability and flexibility to calculate and adjust the takt to changing demand).
Scheduling that way makes sure that things will flow and WIP doesn’t build up on the line but you’ll have to accept that there might be long idle times if the raw processing times vary widely from one station to the next. This is due to the fact that you should wait out the takt time before you move the product forward to the next station. That kind of thing is hard to execute on the shop floor as workers don’t want to look idle. Management usually is also more concerned with high utilization and high efficiency as they are concerned with flow. But flow – even though slow flow – produces at much higher rates than high utilization of workers and work centers. This is simply due to the fact that when you have high utilization with variability than WIP will build up between the station. And the more WIP you have the longer your cycle times will be and the lower your output rate.
It’s not very intuitive but if you run a line slowly with flow, it will actually run much faster than you think.