Accelerating Innovation in the Government-Industry Relationship

Part-Centric versus Drawing-Centric Operational Policy

Imagine rows and rows of drafting tables, in a vacuous warehouse-like work space, cigarette smoke billowing into the collective air of hundreds, sometimes thousands, of engineers and drafters. This is the birth place of the drawing tree as the definition of a product, and where the drawing tree and its constituent drawings became the centerpiece of configuration baselines for decades. Interestingly, since the initial ISO10303 STEP initiative in the 1980’s, the international community embraced the notion that parts and their relationships, not drawings and their relationships, should form the backbone of product definition. However, old “best practices” die hard, and even today, many Government-Industry contract CDRLs call out the drawing tree as the configuration baseline! Rather than belabor why the drawing tree as a product baseline has devolved from a “best practice” to a “worst practice”, this segment extolls the virtues of product definition via part-to-part relationships in a product structure.

So what is the state-of the-art? In the commercial realm, all ERP implementations are part centric; all major PLM implementations are part centric; all major MRO implementations are part centric. Part centric is the new “best practice”. Drawings are now looked at as follows:

  • A drawing is a graphic depiction of a part or set of parts
  • A drawing tree is a hierarchy of those graphics depictions
  • How you organize your drawings are dependent on such ideas as assembly drawings, tabular drawings, mono detail drawings
  • Specifications on the drawing are often not machine readable for query
  • Parts lists are better expressed in a product structure not on the face of a drawing

Although parts are real world touchable items, in information systems, parts take on an additional abstract role. They are the item of interest which you can describe in any numbers of ways – drawings, specifications, 3D models, associated attribute data – and assemble with other parts via relationships. All of these representations and shape definitions of the part are limitless, and even then, have no impact on the existence of the part. With other relationships you can connect parts to contracts, suppliers, configuration items, change orders and similar objects.

The requirement that a part needs to potentially assemble with other parts presents the most compelling argument for a part centric view. Assemblies, after all, are human abstractions for the conceptual convenience of a person’s discipline. Engineering generally organizes their assemblies into a functional breakdown such as structural, hydraulic, electrical, and propulsion. More often than not, manufacturing organizes their assemblies into a building sequence where portions of the structural, hydraulic, electrical and propulsion systems may all exist in a single assembly that, by the way, does not even exist in the engineering assembly. And the MRO discipline, often generates their own assemblies as a hybrid of both. Quality, rightfully so, insists that all the components of all these different views of the assemblies reconcile unless in production the differences are supported by deviations, waivers or a re-interpretation of consumables like adhesives and lubricant. The term product structure encompasses the net set of these assemblies. The term Bill of Materials also refers to the structure of these assemblies. The views are often referred to as the multiple view of the Bill of Materials. This is where terms like EBOM, MBOM, SBOM originate. An interesting fact is that in an information management system, the same abstract part, can be “re-related” in many different assemblies and views of assemblies without ever having to change its graphic depiction or geometric representation.

As a non-exhaustive list, the part centric approach provides operational efficiencies such as reconciliation of “As Designed” versus “As Delivered” product; flexible queries such as “where used”; improved obsolescence planning and execution; explicit definition of alternate and substitute parts and where they are used; and better tracking of the actual configuration of product instances and configurations items.

For some, especially those already in the part centric school of thought, this segment may be considered somewhat simplistic, but does even then highlight a topic of controversial interest – single BOM vs multi-BOM implementations. For those still in the drawing centric school of thought, this segment is probably heretical. Interestingly, often an industry partner is already part centric and has to incur additional burden just to present a drawing centric baseline as a CDRL delivery. In this modern day, the entire product structure with associated drawings, and ideally CAD/CAM/CAx models, should be transferred in a machine readable format between the industry and government information systems. This is where work statements need to evolve.

Based on striving for increased operational efficiency and flexibility, becoming part centric is a critical action for accelerating innovation in the Government-Industry relationship.

Back to Accelerating Innovation in the Government-Industry Relationship -- Introduction