In Episode 76 of Razorleaf’s Stay Sharp podcast, Decoding Product Structures in PLM, co-hosts Jonathan Scott and Juliann Grant are once again joined by Razorleaf UK’s General Manager, Michael Welti. Because Michael has worked on many of Razorleaf’s migration and integration projects, he brings to the table a wealth of knowledge about data models across PLM systems—in particular, Dassault Systèmes. In this discussion, the trio examines what data models are, why they might change, and how changes might impact you if you’re doing PLM.
Why are data models important?
Data models are a blueprint or representation of data—or in Michael’s terms, a definition of objects and links—that organize and standardize how data is stored and retrieved. What’s important for this conversation is that is that an application’s data model characterizes how the application works, structures data, stores it, and more. “The record types, the relationship types, what’s allowed to be connected between different relationships,” Jonathan says. “That’s what we’re getting at with data models.” What’s more, each application has a different approach and model, which significantly increases the challenges of integrating different applications throughout your enterprise.
Jonathan brings up the idea of attributes as a part of a data model, and Michael agrees that they’re how you can manage things like a bill of materials (BOM), noting “The things like revision schemes and lifecycle policies that go into how you approve and release and change those BOM parts as part of a process. [It’s] sometimes quite different to how you do CAD files.” Attributes are also the kind of information you need in your ERP system, he says, “Part numbers, descriptions, classifications, anything that you end up needing to help control that on the ERP system. And that usually helps define and drive what those attributes in your parts are.”
While Michael, Jonathan, and Juliann aren’t saying you need an in-depth understanding of every variation of data models—or even the model in your own system—they are saying you need to thoroughly understand your own data. The capabilities that the PLM, or any other enterprise platform, offers “are rooted in the data model,” Jonathan explains. “So think how it delivers the capabilities or if it delivers the capabilities you want, based on how it represents the data.” Michael agrees, saying, “Understanding your own data is the best place to start. If you’re thinking of getting a new PLM system, get a good feel of what exactly you want to be managing.”
Why change a data model?
One fact about software and technology is that they always evolve. A case in point is Dassault Systèmes, which is currently changing their data model. Jonathan offers his opinion on why: “It’s going to be easier, a better representation of the data in the real world if we offer this new data model and these new capabilities around it.” Moreover, the trio agrees that Dassault’s new approach—which takes multiple tools with multiple different data models and brings everything together under one consistent schema or data model, known as the unified product structure (UPS)—will be critical for the next generation of PLM and the emphasis manufacturers are placing on creating digital threads and digital twins. Michael is intrigued. “It lets you do some very interesting things. It means your engineering BOM could be made up of multiple CADs. You’re working on one BOM structure, but one part is a SOLIDWORKS part, then the other part is CATIA. You might have a bit of software parts,” he says. “It’s all in there, all interlinked, because it’s the same data model and schema. It can all interrelate and be seen as one unified product.”
Michael explains that all of the PLM systems are working to do the same thing, to “represent more and more of the wider thread in one platform.” They’re all responding to the greater demands on their products—such as the digital thread—and realizing their existing data models aren’t doing what they need them to do. Or potentially, Jonathan notes, they’re taking a future view and realizing they won’t do what will be necessary in the future, so it’s time to start changing them now. The web of the digital thread is getting much broader, Michael says, “And the various vendors are trying to provide tools and technologies to expand what you can do.”
Learn More About Data Models
The full podcast contains more discussion and details on a variety of topics, including how new data models could enable better collaboration among teams, how to evaluate the tradeoff between cutting-edge technology and openness in your PLM system, what you need to do about data model changes, what your starting point for PLM platform decisions should be, and more.
Check out the full conversation in Stay Sharp Episode 76: Decoding Product Structures in PLM, and join us each week for a new podcast.
Follow the Razorleaf Podcast, Stay Sharp in Digital Engineering on:
