Team and Product Data Management (PDM)Thus far in the Associativity and Part History sections, we've assumed that we have one person at one workstation in one data workspace doing the entire job, in the mode of Leonardo da Vinci (i.e. not very realistic). More likely we have a team of people (a few, or dozens, or hundreds) that need to contribute to the activity, perhaps even distributed over company boundaries. Let's look at our simple mouse example and now distribute the various tasks to Dave Design, our customer's industrial designer; Ed Engineering, our mechanical engineer; Matt Manufacturing, our mold tooling specialist; and Nick NC, an part programmer at our machine shop subcontractor. For sanity, we'll assume that all are using the same CAD/CAM software system (more in the Imported Data section) and that they are all on the same local area network. Dave begins to work on "version 1" of his sculpted mouse concept. Ed, Matt, and Nick might be able to get a jump on things, even before they get anything from Dave, by working on related elements of their activity (Ed on the rib/boss, Matt on tentatively selecting a mold base, Nick on fixture design).
When he's ready (earlier rather than later, we hope), Dave releases "version 1" of the "industrial design" model to the team. Dave then begins to contemplate and investigate additional refinements in the direction of an eventual "version 2", and therefore Dave retains "ownership" of the industrial design model. Ed "references" (if he has the "authority" to do so) the industrial design model as the "parent" for his engineering model. He adds additional features such as the rib and boss, and at some point releases "version 1" of his engineering model. Ed retains ownership of the engineering model. Matt references the engineering model as the parent of his mold base model. He defines parting surfaces, splits the mold into core and cavity, and details the core and cavity (e.g. thru-holes for locator pins). So far, so good. But now, Matt sees that the rib is too narrow to be molded properly... so he asks Ed to widen the rib (or asks Ed to relinquish ownership so that he, Matt, can edit the rib himself). The engineering model is regenerated with this rib change and released as "version 1a" of the engineering model. Matt's mold base model detects a change to its parent so it regenerates, and then the mold core and mold cavity detect a change to their parent (the mold base) and they regenerate. Matt likes the result and releases "version 1" (or he may choose to call it version 1a for consistency) of his mold core. At this point, Matt may also generate a bill-of-material and prepare a traditional mechanical drawing of the mold core and release these items to purchasing and process planning. Nick references the mold core model as the basis of his fixture design and NC programming. He gets a couple clamps selected and has a few NC operations defined when... ...as we might have known, marketing changes its mind and wants the mouse to be wider. Dave complies and also domes the top sculpted surface a bit more. He releases "version 2" to the team. In a utopian world, everyone likes the new, version 2 design and allows their "children" models to regenerate (successfully we hope) to reflect the ripple changes up the parent/child chain. But more likely, things don't proceed in a linear fashion and instead design recommendations get from time-to-time "bounced back" to their originators (perhaps Nick back to Matt, Matt back to Ed, or even Nick back to Dave). In fact, in a well-run production use of team CAD/CAM this is exactly what we want to happen... downstream issues with designs should be raised and resolved as early as possible. The more interchange, the more flux, the earlier... the better. This is in fact exactly what we're after with "concurrent engineering"... the ability for multiple people to be working on multiple interrelated data sets at the same time... and when properly implemented is perhaps the key to achieving significant reductions in time-to-market. But this kind of desirable interchange and flux in the context of linked, associative models can only work when deployed in conjunction with a good "chaos manager"... i.e. an effective, pragmatic team and product data management (PDM) facility that is tightly linked to the CAD/CAM system. Imagine how much we need to manage the chaos when we have a dozen or more people working on hundreds or thousands of piece parts, each part undergoing a good handful revisions. Such a PDM facility provides at least the following mechanisms:
These minimum PDM functionalities are often directly bundled into and integral with the CAD/CAM system and might be referred to as "team data management (TDM)". Full-function standalone PDM systems go further to also include such advanced functions as:
CAD/CAM systems are delivered with PDM in different ways:
|
|
|