Last week in Part I, we discussed the nature of CAD data and the challenges of managing this data and the engineers who create it.  Balancing between the need for freedom to foster creativity and productivity versus the discipline required to convert this output into tangible results is daunting.  Creating systems that accomplish this end goal is no small feat.  We will dive into how these systems try and address these challenges and the compromises companies must make when choosing the best approach to harnessing their engineering output.

When you vault CAD binaries in a Product Lifecycle Management (PLM) system, the stakes rise immensely. The PLM must have a data model that can capture the design product structure. The architecture must be able to support the transmittal of large amounts of information. The security must ensure that the information is locked up tight to avoid theft or error. All of these things impact the user experience from an interface and speed perspective. On the other hand, not having some sort of engineering data in your PLM can severely cripple the benefit of having PLM in the first place. Many companies avoid the whole issue by adopting a user friendly PLM that either has no ability to capture CAD information or allows adoption without engineering. I am sure these systems yield significant benefit to the companies or they wouldn't bother putting them in place but they are missing a huge value opportunity by excluding engineering. In general, having everyone connected to the same system with a single data vault is preferable. This ensures early access to engineering which allows for optimization of cost and delivery. Change management becomes a much more integrated function and change is obviously one of the major drivers of product cost and delays. Early access minimizes late change and having engineering integrated into the change process gives you the ability to respond more rapidly. Some of this can be done without having engineering data in the system but it requires a lot of manual work to get product structure synced up between engineering and the rest of the company. 

I think we can all agree that having engineering integrated in with the rest of the enterprise via PLM is a desirable goal. The question becomes how can a PLM system be robust enough to meet the needs of engineering without saddling the rest of the organization with poor performance and a daunting user interface? To link it back to my original analogy, do we choose the whirlpool that will take the entire company down or do we steer towards the rocks and the scary man-eating monster? We will explore a couple of solutions from Oracle that can possibly give companies the best of both worlds. Oracle offers an Engineering Collaboration module for most major CAD tools that allows engineering data to exist in their Agile PLM environment but as a separate design object. This gives engineers the freedom to control access to the information by only publishing out what they want the rest of the company to see. It is also a fairly open environment that gives them the opportunity to iterate if they choose. Unfortunately, as I mentioned earlier, CAD tools are very complex and sometimes third party solutions fail to meet the engineer's needs. In these cases it is possible to allow the engineering information that we term "work in process" (WIP) to stay in the authoring tool's product data management (PDM) application and use connectors to publish over to the PLM environment based on state changes like release or another type of promotion. This indirect approach allows the engineers to use native tools and keep their environment relatively unchanged but gives the enterprise visibility when appropriate and doesn't burden the enterprise with overly sophisticated or complex interfaces and large data sets that can impact performance. We also have a webinar series that goes over the different types of integration. You can watch the recordings here.

Please check out our Engineering Integration/Collaboration Hub or read our Engineering Collaboration e-Book.

[Edit: Updated and reposted; original post 2015]