DReimer
Mechanical
- May 20, 2005
- 55
We are a manufacturer of industrial machinery. Our machinery comes in many different optional configurations, with occasional customization (10-20% of machines).
Our bills of material have traditionally been structured along functional areas. An engineer would then select a bill for each functional area from a "menu" of choices to match the required configuration of the machine.
Because of the multiple configurations, however, we end up with numerous bills of material for each functional area, representing all the different combinations of options that affect that area.
The net result is that our top level BOM has 50+ items it, many of which are selected from 3-6 choices. As you can imagine, it is easy to make a configuration error, and they happen with unfortunate frequency. It doesn't help our production process if a major subassembly is missing when the parts hit the floor. My department (engineering) takes a great deal of flack when this happens.
Add to that the clerical burden of maintaining so many BOMs (most of which are phantoms, and differ by only a few parts out of many) and the entire structure isn't very efficient.
To remedy the situation I've gone through a bill modularizing process for one of our machine models, and developed our ERP's configurator to configure the BOM for us. (We've actually used the configurator off and on for many years, but always using our traditional bill structure which made it very difficult to maintain the configurator and evolve it with the machine).
The modularized bills consist of a common parts phantom bill and a bunch of bills that add parts for each option. So far so good. We've tested the new configurator and it seems to work very well. I've documented it extensively (our MRP system is Unix based and very non-friendly, so I've "mapped" the bill structure in Excel). The modularized bom will also facilitate a lean manufacturing intiative should be get around to launching it.
Here is my problem:
We supply a parts manual with every machine. The previous bill structure facilitated this as we could generally get assembly drawings that matched the MRP bill of material. We could then export the bill to Excel and create a parts list. We were reasonably certain that the parts list would be accurate, because it came directly from the MRP database maintained with an ECO system.
Now, however, the bills of material used to build the machine are all over the place. There is no reasonable way to make them match up to a drawing. This means that a parts list for any drawing we use will need to be maintained separately from the database. Talk about potential for errors!
Here is where the "different masters" comes in. Our parts and service department is generally happy with a simplified parts list that identifies parts that would conceivably need to be replaced in the field. But our manufacturing department wants complete parts lists in a maintained binder so that they can identify "missing" parts.
Both of these require that the parts list be accurate (our parts guy has been known to sell $4000 subassemblies based on the parts manual without checking the actual bill of material. His belief, I guess, is that every configuration of every subassembly should have its own drawing with the correct part number). As mentioned, because we are now trying to maintain two bills of material for the same machine, maintaining this accuracy will be difficult.
How are other companies dealing with this? I posed the question to a consultant that did a 5S exercise here, and he says that in a modular bill/lean environment companies use photographs and work instructions for internal use. He had no answer for the parts manual question.
Any stories, advice, tips would be appreciated.
Cheers,
Dean
Our bills of material have traditionally been structured along functional areas. An engineer would then select a bill for each functional area from a "menu" of choices to match the required configuration of the machine.
Because of the multiple configurations, however, we end up with numerous bills of material for each functional area, representing all the different combinations of options that affect that area.
The net result is that our top level BOM has 50+ items it, many of which are selected from 3-6 choices. As you can imagine, it is easy to make a configuration error, and they happen with unfortunate frequency. It doesn't help our production process if a major subassembly is missing when the parts hit the floor. My department (engineering) takes a great deal of flack when this happens.
Add to that the clerical burden of maintaining so many BOMs (most of which are phantoms, and differ by only a few parts out of many) and the entire structure isn't very efficient.
To remedy the situation I've gone through a bill modularizing process for one of our machine models, and developed our ERP's configurator to configure the BOM for us. (We've actually used the configurator off and on for many years, but always using our traditional bill structure which made it very difficult to maintain the configurator and evolve it with the machine).
The modularized bills consist of a common parts phantom bill and a bunch of bills that add parts for each option. So far so good. We've tested the new configurator and it seems to work very well. I've documented it extensively (our MRP system is Unix based and very non-friendly, so I've "mapped" the bill structure in Excel). The modularized bom will also facilitate a lean manufacturing intiative should be get around to launching it.
Here is my problem:
We supply a parts manual with every machine. The previous bill structure facilitated this as we could generally get assembly drawings that matched the MRP bill of material. We could then export the bill to Excel and create a parts list. We were reasonably certain that the parts list would be accurate, because it came directly from the MRP database maintained with an ECO system.
Now, however, the bills of material used to build the machine are all over the place. There is no reasonable way to make them match up to a drawing. This means that a parts list for any drawing we use will need to be maintained separately from the database. Talk about potential for errors!
Here is where the "different masters" comes in. Our parts and service department is generally happy with a simplified parts list that identifies parts that would conceivably need to be replaced in the field. But our manufacturing department wants complete parts lists in a maintained binder so that they can identify "missing" parts.
Both of these require that the parts list be accurate (our parts guy has been known to sell $4000 subassemblies based on the parts manual without checking the actual bill of material. His belief, I guess, is that every configuration of every subassembly should have its own drawing with the correct part number). As mentioned, because we are now trying to maintain two bills of material for the same machine, maintaining this accuracy will be difficult.
How are other companies dealing with this? I posed the question to a consultant that did a 5S exercise here, and he says that in a modular bill/lean environment companies use photographs and work instructions for internal use. He had no answer for the parts manual question.
Any stories, advice, tips would be appreciated.
Cheers,
Dean
