Infor CloudSuite Industrial APS Best Practices #6 – The power of clear and correct instructions
When I was younger, I was a snickerdoodle man. I loved the smell of cinnamon and sugar that wafted up from a warm batch straight out of the oven. But the first time I tried to make these delights on my own, the cookie dough simply melted in the oven, and I ended up with one giant, rubbery, cookie sheet-sized rectangular cookie. My buddy and I tried to choke some of it down, but very soon gave up on that.
What went wrong?
After a furious search, I found I had copied the recipe down incorrectly. I’d missed the baking soda and cream of tartar. In manufacturing terms, I was working off of a bad BOM.
Now imagine if I had attempted to apply my clearly flawed recipe copying process on hundreds of items and even more updates as part of my business.
We know we need accurate BOM data to get good plans with APS. But how do we design a process to ensure that accuracy? Well, we can start by taking the principles we learned earlier with cycle counting and applying them here. In this article, Jim Black, former director of consulting at Infor, tells us exactly how to do that.
How to make sure routings and BOMs are accurate
Look closely at the CloudSuite Industrial (SyteLine) documentation, and you’ll notice that Infor uses the acronym BOM to mean Bills of Manufacture and not just Bills of Material. APICS doesn’t define it that way, but I like that it encompasses both the material and process-related aspects of the product structure. In fact, the process industry often refers to it as the recipe, which of course includes both input materials and “cooking” instructions.
In any event, the point here is that the data used for planning needs to be accurate. But how do you measure the accuracy of a BOM? There’s a lot of talk about the need for accuracy, but seldom (if ever) do you see a specific approach to determine it—until now.
Many years ago, I created what I refer to as a BOM Accuracy Report. It was designed to be easy to complete yet effective to use. It is specific to CloudSuite Industrial, but the concept is applicable to any ERP product. Here’s an example.
For materials, you’ll audit to verify that the:
- Right items are listed
- Required items are included
- Quantity per amounts are correct
For operations, you’ll audit to verify that the:
- Right resource groups are listed
- Production rate is reasonable
- Operations are in the right sequence
- Operations make sense
- Control points have been set correctly
The process requires an auditor to pull at least one job order at random from the floor (regardless of stage of completion) for auditing. This should happen at least once every workday. And the auditor should avoid choosing a job for an item that has already been audited this month.
With the job in hand, the auditor will record the output item number of the job (notice that the job number itself need not be recorded). He or she will then examine each of the listed input materials to determine that (a) each listed material is correct for the output Item, (b) all the necessary input materials are included, and (c) that the Quantity Per listing is correct for each material input. Each test will receive either a Yes or No score in its respective column. Yes, it’s a binary option with materials. It’s either right or wrong.
Let’s see how this would work. Let’s say that to make the output item (a men’s bicycle), the job shows the inputs to be:
- Frame (women)
- Chain assembly
In this example, the auditor finds the wrong frame is listed so enters an N in the “Right Items” column. The auditor also finds no seat listed (and it should have been) and so enters an N in the “All Items Included” column. Finally, the auditor finds the Quantity Per for all 5 listed materials was correct except for wheels – it showed 1 instead of 2. The auditor would, therefore, mark the “Quantity Per” column with an N as well. Although 4 of the 5 were correct, the column is given an N. You may think that is a rather severe score, but that’s the intent. Do not modify the form to allow partial scores like 80%. You’ll only be fooling yourself, and you’ll certainly add significant complexity to the scoring process. The fact is that a particular job order could not be built because of these material BOM errors. This is what you’re auditing for.
Now let’s move to the operations portion of the BOM (the routings). We can be less absolute here. First, we check that the resource groups are correct for each operation. If anyone is wrong, the auditor enters an N in the “Right Resource Groups” column. We then check if each operation specifies a reasonable production rate. Now, this is where we have some latitude. Close counts here. It’s a matter of accepting the value shown as reasonable or not – and reasonable to reality, the demonstrated rate, not to a standard rate.
The next three checks—Right Sequence, Right Operations, and Control Point—are optional. Sometimes these have tangible bearings on the plan. Many times they don’t. For example, if you’re backflushing material, then the control point is more important. If you’re not backflushing, who cares? The key is that if you include them in the audit, then they are given a Y or N, and their score must be considered just as any of the other columns. No freebies, folks. That helps no one.
Now a little note on “Right Operations.” When checking for this, you’re determining whether what you have modeled in the BOM is necessary. For example, just because a product goes from milling to drilling to grinding doesn’t justify having all those steps as operations in the routing. Instead, it may be highly desirable to treat those operations as a cell. You’d then model that single cell as the operation (e.g. Fabrication Cell) in the BOM and assign it a reasonable production rate. In other cases, it may make sense to completely ignore some physical steps. Should QC be an Operation? Proper BOM construction is a topic for another day, but this is what you’re checking for.
With all of the operation scores, they’re either Y or N. Again, this is a severe test, but that’s the intent. Do not modify the form to allow partial scores. As with materials, you’ll only be fooling yourself, and you’ll certainly add a lot of complexity to the scoring process. Keep it simple, and you’ll keep it effective.
You may have noticed that this audit only covers job BOMs and not current BOMs. If so, kudos to you. Given that a very high percentage of companies using the software copy the current BOM to the job BOM, this is of little concern. Nevertheless, bear in mind that any errors you find will probably require that you modify the current BOM.
The auditor does this each working day of the month, and at the end of the month, computes the aggregate results. In our example, we audited a total of 20 BOMs (by the way, no two for the same item). Of those, only 15 scored ‘Yes’ in every column. We can then report to management that our BOM accuracy for the month was 75% (15 of 20).
The key, of course, is investigating the problems and taking corrective action. If that is done, then the score is only going to improve over time. In fact, the score sheet helps management to focus on which area is causing most of the problems. In our sample, it appears that we have problems with the “Right Items” and “All Items Included” criteria. The BOM Accuracy Report, if created in a spreadsheet, should have two tabs: the scoresheet and a graph. The graphical presentation visualizes where the problems are. And it’s helpful to have it scroll every three months.
It is important that the auditor not be the same person who created the BOM; however, it should be someone who knows the manufacturing process very well. They will red-line the selected jobs and work with the appropriate parties to ensure corrective action is taken. The auditor may be a production supervisor, a senior operator, or a QC person. My preference is for a senior operator as they typically bring the most common sense and independence to the process.
By the way, regardless of job order volume, auditing just 20 jobs a month is statistically valid assuming they are selected at random (and that no two jobs are for the same item).
In closing, let me add a few observations. This audit process isn’t perfect, as it doesn’t cover everything.
For example, the process doesn’t test for the proper use of phantoms. Phantoms introduce far too many nuances to lend themselves to such a simple test. To cover those, a separate (and normally one-time) review of phantoms should be undertaken. You’ll notice as well that this process doesn’t attempt to evaluate whether fields such as Move, Queue, Setup, or Finish on the operations form are properly maintained either. In most instances, however, these fields should not be used. They normally add more heat than light to the BOM. But, as with Phantoms, a one-time review of these fields should be made before you go live and then perhaps once a year thereafter.
So while the process isn’t perfect, it is excellent. Not only will it provide insight as to how much confidence you can have in the plan that’s created using your BOMs, it will also help you improve your BOM accuracy over time.
You want to be a company your customers love to do business with—one that’s reliable, that makes and keeps its promises. A great planning process helps you do that. But you can’t make bricks without straw. The straw, for planning, is good data.
There are six practices associated with making sure we feed APS accurate information.
- Record labor and material transactions timely and accurately
- Keep inventory records accurate
- Close job orders in a timely manner
- Keep routings and BOMs up to date
- Give purchased items accurate lead times
- Depict work center capacity realistically
We’ve covered the first four of those practices. In the next post, we’ll look at some critical tips on how to keep your purchase part lead times up to date.
Learn how to use APS to improve your on-time shipment rates
To plan with APS, you need to know how it works. The best way to do that is to take the self-directed APS course on Infor Campus. In this hands-on class, you learn the seven steps of planning with APS plus all the key forms and fields.
About the author
John Brown | Education Product Manager, Infor
John is the Infor Education program manager for CloudSuite Industrial, CloudSuite Business, and CloudSuite Industrial (SyteLine). He joined the Education department in 1999, back in the Symix days.
- Education & Transformation Services
- Industrial Manufacturing