Hello everyone,

Any idea on how to make all items in V6 database unique automatically without user actions? In V6 default database Part Number must be unique, otherwise item can not be saved to system. But this Part number value is user editable => so users have to manually take care that all parts are uniquely named before save. Naturally this is just impossilbe if we think proper usability.

So there must be a mechanism that can be set to create automatically unique identifiers that are not user editable, so that saving any item to system creates automatically a new Item ID from a sequence defined.

I don't know if Part Number is hardcoded to be unique or not. If it is , then it must be made not editable by user => how to do that? Didn't find instructions for this from online docs or other V6 resources, I bet that someone has done this already, so can you shed some light on topic?

 

 

 

 

Yes, part numbers should be unique, but long experience tells me that they should remain editable by the user until first release. Most engineers are OCD, and hate skipping numbers. They also really like to have their part numbers grouped intelligently. I don't know why, because that logic is typically gone after about the third or forth iteration. But, if you don't allow them to tweak their part number prior to release, they will scream.

We had a very difficult time moving our tooling group to V5. Their typical process didn't have them assigning part numbers until after the design was complete - their part numbers included the page number of the drawing in the number, and they didn't know what page number a part would be on until they completed the drawing.

Hi Jim,

Hmm... okay, I admit; this can be process related. In some processes - typically if PLM system is not actually the Master source for component coding - Part number needs to be manually editable but still unique. But still a question remains; code (or part number) must come from somewhere, it's not invented by user alone. So I am looking for the source of this code or number ;it should come from sequence of codes. Is there any native capability in V6 that generates such a code? (Like Enovia V4 COMPID generation or SmarTeam Sequence designer)

I have seen a lot of silly things built based on codes; typically 'descriping code' tends to be a source of unhealthy processes; rules how coding happens change when processes change. So it seems that code must be really a simple code from single sequence, without any coding of department, part type, even drawing size (seen that also!) written to the code. Just plain number/lettercode that identifies unique item is healthy base to build on.

Typically, in a large companies where VPM or VPLM are used, users can not create part numbers in the database.

The CM department (configuration management) is responsible for the part number allocation and creation. Usually, some internal application is used to create the PN according to the company's requirements.

The user then searches the DB for the part number(s) allocated to him/her.
The part is then loaded into CATIA, modified and saved back to the DB.

Hi Andre,

Yes I know, I have been working in two gigant size companies that are leaders in their industry on enhancing these things. Horrible process isn't it? Basicly this is a process that can/needs to be changed to more efficient.

Even when CM department (or similar, this Really depends on process) creates this code, it just not come from air; there is a sequence and logic that creates it. That logic must be connected to part number creation process, so that original part can be created fast and efficient. =>still searching for V6 capabilities for this, no matter what is detailed process, there still a need for this.

I don't know about V6, but that functionality is not included in VPLM.

I agree that this process is very cumbersome:

Company has to define rules to allocate PN.
Company has to create software to create and manage PN.
Company has to create another app to create Parts in VPLM... expensive
User: curse, curse, curse...

I agree that there should be something in DS products to do this. I keep looking.

Some of this is getting into the differences between Development and Production. While still in the Development phase of a product, we have much greater flexibility.

While we have rules and applications written around the basic "Drawing Number" (this is the base portion of the Part Number), we allow the Designer to assign the "Dash Number" (this is the portion of the Part Number that comes after the "Drawing Number) of the part. During development, the designer is constantly adding and deleting parts from his design. Sometimes this is because they are looking at manufacturing constraints, other times it is because they are looking at weight, strength, and durability constraints. VPM and CATIA V5 have forced the designer to create part numbers before the design is final, so they check these "dash numbers" out in sequence. Then, when the design is complete, they go back and re-assign their "dash numbers" so that they don't have any gaps in the sequence, and so that they have some level of intelligence (for example, maybe all assemblies have to have the "lower dash numbers").

After the design has been released, the flexibility in part numbering is greatly diminished. Typically at that point, the designer just assigns the next available "dash number" to the part. At this point, the intelligence in the part numbering is completely lost (in the above example, if you need to assign a new assembly part number, you cannot give it a lower number, as they are already completely used).

Forcing the designer to go to CM ever time he decides that he needs a new support bracket, or he needs to make his assembly from 3 parts instead of 2, would really slow down the system.

Jim,

Now we go to process details; I see clear similarities in your process to those processes I have seen. In case where I have worked we have had a successfully flexible system with roughly following process:
- There are two separate attributes that affect on part identification
- First one is "System ID" => automatically generated ID, that user can not affect to. Just a stupid code, but unique to system.
- Second one is actual company Item ID (closer to part number in this discussion) In the beginning of process (development time) this is empty; gives total freedom for fast modifications. Later before release and production time this is filled with company item ID.

=> simple as that.

Anyway; this is not the thing that can be done in V6 default environment, there is no system ID and V6 is forcing Part Number to unique keeping it still manually managed. In practise this ould mean that in beginning of process users would have to fabricate their own fake codes that are somehow unique. I don't want that.

Where I want to go is that we really take a fresh start and use system automatically generated ID (which doesn't exist in default V6) as master, so that user has no manual possibility to affect on it. Then later downstream processes would just follow this system code. Of course several codes would not ever be used, but who cares; they are just numbers.

Anyway; I don't want to go to process details, just interested if anyone has found a way to ensure Part Number uniqueness woithout manual interaction within V6 environment.

I agree with your points.

As for the manual interaction, can it be done with a User Exit or other customization? If not, what about an external application? We use a web page that is connected to VPM that allows the user to create part numbers and connect them properly to the appropriate meta-data. It also allows them to create new CATParts and CATDrawings based upon the correct seed models. This keeps the part numbers unique as long as the user uses the web page. Of course, they still have the ability to use standard VPM functions, in which case management is still manual (a user exit does exist to ensure uniqueness, but none of the other things the web page enforces).

Hello Jim,

As getting a code from some sort of sequence is a need that literally EVERY customer has; I would have assumed that there would have been some sort of proposed solution or configurable way of doing this. Even if here would have been such a solution, there would have been a need for user exit type of customisation, since many companies would not settle to standard solution, but would build really customized solution. (which I personally see at a point of changing system a bit silly; fresh start is what many systems need, settling to historical coding principles is not healthy - but this is just an opinion, not what will maybe happen in reality) But still... there are many companies who don't have resources or will to develop their own solutions, but would willingly settle for a good proposed standard solution.

As I mentioned at one other Post (general discussion - future of PLM, started by Tim of Boeing) I find it weird that even basic processes, like getting a code, are not proposed in default system, but really left under construction, trusting that all companies will create their own solution, is definetly Not good. I think that many companies that have taken system customization to high level have noticed that standard solution is in the long run always better than customized solution. When new toy like V6 appears to market, I think I am not only one who dreams of getting many traditionally customized features to be replaced with standard code. That has several benefits which should have seen as competitive benefits by PLM vendors. Now when many companies are in evaluation phase towards V6, DS should come up with as good as possible solutions for existing processes.

Traditional approach of DS has been that they don't propose much for processes, but merely try to create solution based on some specific customers processes. Now when they have a long history of process development with several big companies; I think they should have a capability to come up with proposed start to end solutions on top of V6.

...it sems that I got carried away. Again.

Another thing related to code's and part numbers in V6. Has anyone found a user exit (or system configurable) functionality to affect on Duplicate -command?

I find this Duplicate as a good functionality in principle, but standard capability that it offers on changing attributes when duplicate is initiated is rather limited. It only allows to add a Prefix on Part Number. In most processes where Part Number is used as company Item ID, then adding prefixes on it is seriously No-no => Part number should be changed to different Item, if item is duplicated.

In my mind functionality (for generic template based project type process) should be following: (I know that this is common process for many industries, but naming can be different in different companies, so I explain basics of process here to make sure that we discuss about same things)

- If Item is duplicated Part number should be assigned from company specific sequence (as in new part creation in previous posts)
- If Item is duplicated an attribute that defines Item role should be switched from "Template Item" to "Project Item"
- Logic that blocks duplication of "Standard Item" role parts, only "Template Parts" are duplicated


Short explanation about process that I am looking for to be supported:

There are three Item roles:
- Template Item = Item that can no be directly used in production, typically a parametric template pat or assembly.
- Standard Item = Item that is same from project to project, typically purchased standard part, bolts, nuts etc.
- Project Item = Project specifically designed item, can be based on Template Item or designed from scratch, or copied from other project and modified to fit to this project.

Typical Template Item assembly includes both Standard Items and Template Items. Standard Item assembly can include only Standard Items.

When a Project is initiated, first task is to find appropriate Template assemblies and Duplicate their content in selective way, so that Template Items are duplicated, but Standard Items are not. Template items become Project Items, an they are assigned with new Part numbers. Project Item assemblies are then those, that go to actual project specific design and finally are manufactured. This kind of process can very well take in use advantages of knowledge capabilities of CATIA, since developing Template Items to intelligently changing Items is basic advantage of such a process.

==> I see that Duplicate function can ALMOST do this "move from Template to Prject specific" task, but needs certain configuration to really fulfill the need. I went through the online docs, but didn't find anything from there for modifying Duplicate functionality; I wonder if anyone has seen any capabilities that can be added on top of such a V6 functionalities? I bet that there is certain level of openness in these V6 functionalities, but I don't know (yet...) how to customize these capabilities. I would really hope that this is NOT CAA development, but rather a simple system configuration task with sensible UI.