In the movie LA Confidential there are many surprising plot twists. I was very surprised to learn these as the movie unfolded. I never expected that the chief was in on it nor did Kevin Spacey who took a bullet for his misjudgment. There are plot twists and surprises in PLM as well but hopefully we don't have to take a bullet to learn them. As we know three of the major Product Lifecycle Management (PLM) vendors originally were computer aided design (CAD) vendors and their PLM solutions evolved from their need to manage the complex relationships between the various files generated by their CAD applications. As the story goes it makes perfect sense to take the engineering data created in these CAD systems and then leverage it for creating Bills of Materials (BOMS) for configuration management which can then be pushed on to enterprise resource planning (ERP) tools. Ideally, you have a single system of truth and the product design kicks off a process that ultimately yields the manufacturing information needed to build the product. This sounds very logical and if you are not intimately familiar with the business practices and design processes of most companies you would probably think this is the norm when it comes to how most companies use PLM. A recent article by Alex Gammelgard on the Arena Website poses the fair question, "Why is it so hard to move a BOM from engineering to manufacturing?" In his article Alex explores the differences between manufacturing and engineering but really only skims the surface of the fundamental issues that make it very challenging for having an automatic handoff from engineering to manufacturing. The reality is that very few companies use their CAD data to drive their manufacturing BOM. This article will explore some of the challenges around this approach and what steps companies need to take to accomplish this desirable objective.
The fundamental issue that prevents most companies from sharing engineering information with the rest of the company is that the format and content from engineering is ill suited for creating a bill of materials. I have had numerous conversations with companies that were dismayed to learn that their engineering organizations have taken the most expeditious paths in creating their designs which often means that the structure of their assemblies and the content of these assemblies does not reflect the requirements that a configuration manager or a manufacturing engineer would need to successfully leverage this information. Sometimes the structure is close and companies will use the engineering information as the basis for the manufacturing BOM (MBOM) but the way a system deals with this issue can make this approach problematic especially when iterations are still being done on the design itself. From an engineering perspective it doesn't make sense to include every fastener or component in the CAD model since they may not be essential to the design. It would be very time consuming to do this and not really accomplish much from an engineering perspective. From a manufacturing perspective this is a big deal as they need all items on the BOM. Beyond just library components the way an engineer structures their assemblies is based on functionality and has very little to do with manufacturing. The hierarchy and structure may need to be very different on the manufacturing side. Also in many companies engineers don't really worry too much about part numbers and other manufacturing issues when they design. They cut and paste at will from different assemblies to speed up the creation of new designs but from a configuration management perspective this approach can create havoc. The primary issue is that engineering has typically evolved in a context where they were isolated and that the hand-off to manufacturing was manual so their methodologies are not compatible with the current PLM process map.
Process is a problem but the technology itself also creates issues. Even though PLM software is becoming more and more prevalent we still see a large number of CAD only data management solutions around. SolidWorks Enterprise Product Data Management (EPDM) is growing rapidly and I feel certain that more companies are using Intralink for CAD data management than PDMLink for enterprise product lifecycle management. I am not as familiar with the TeamCenter and Enovia market but I would suspect that stand alone CAD data management is more prevalent. Why are companies not moving over in great numbers to managing their CAD data in PLM? I would contend that the requirements for managing CAD data do not translate into the enterprise PLM application. Standalone CAD data management tools allow a certain amount of freedom for the engineering organization that are probably good for them but not for the entire company. PLM enforces a discipline on the engineering organization that may hinder their productivity and may also lack certain features that they use with dedicated CAD PDM. This is why many Oracle Agile PLM clients don't use the engineering collaboration module and why so few of PTC's clients use PDMLInk to manage both CAD and enterprise data. PLM vendors need to really look at this issue and provide better tools and processes for reconciling the differences between engineering and manufacturing. Oracle's design object helps provide a middle ground but they still lack straightforward ways to address the difference between the two environments. PDMLInk burdens enterprise users with an overly complex interface with slow response due to the background commitment to CAD. The irony is that if you looked at PDMlink or any of the other CAD based PLM tools as just stand alone PLM systems they would fair very poorly against the pure PLM systems yet very few companies are able to use these systems as originally expected due to the engineering data structure.
If a company does desire to consolidate their environments and streamline their communications by fully integrating engineering into their PLM environment they need to look beyond the functionality of their PLM tool and really analyze their engineering methodologies. These companies may need to make compromises that will impact their engineering efficiency but will ultimately lead to a more unified environment. Regardless of whether they choose to leverage a single system or integrate between PDM and PLM the data must be adjusted to take into account the need to leverage CAD data for product configuration. Many times the processes engineers use for design were established before this was even a possibility. Moreover the CAD tools themselves lack any functionality to make this process easier. This could be a potential area where a vendor like PTC or Dassault could take steps to differentiate themselves. The ability to create a MBOM straight out of the CAD tool would be highly desirable and would lessen the need for dramatic changes to engineering methodologies.
Having a process that allows product design to drive the product structure from design through manufacturing can be a valuable capability. The improved communication can save companies millions and decrease time to market significantly. The plot twist is that very few companies have their engineering information structured in a way to facilitate this and PLM and CAD applications lack the functionality to effectively address this issue. Hopefully this dirty little secret can be exposed and companies will evaluate their engineering methods in a broader context and PLM and CAD vendors will help their customers out with more robust reconciliation tools. So no longer will this be hush-hush and on the QT.
[Edit: Repost from 2011]
[clear-line]