Leveraging Model-Based Definition with aPriori

Christopher DeRosa: Hello and welcome to aPriori’s 2022 Manufacturing Insights Conference. Today’s expert track will be discussing leveraging model-based definition with aPriori, but first, a little bit about myself. My name is Christopher DeRosa. I’m an application engineer with aPriori, and my background lies primarily in mechanical design engineering as well as CAD and PLM management. So I’m looking forward to today’s presentation and taking you through the different layers of automation we can enable through model-based definition and aPriori’s role in this digital thread.

What is Model-Based Definition in Product Development?

So what is MBD? Some may say it stands for manufacturing builds drawings, which in many cases is true based on traditional 2D workflows. But that is exactly what we are looking to move away from. MBD stands for model-based definition and is the practice of annotating tolerances and PMI data on the 3D CAD model. So MBD serves as this foundation of the digital workflows in design and manufacturing. And by now, most engineering teams are beginning to adopt and engage in MBD practices. And so, the benefits of going to MBD will have a major impact on engineering and manufacturing workflows.

Another acronym we often hear is PMI. So what is PMI? PMI stands for Product Manufacturing Information. This conveys the non-geometric data and attributes that are required to understand the full product definition of a 3D CAD model. So attributes such as your material surface finishes and other testing or inspection processes can all be found in the PMI data. So as we move towards MBD adoption, what does that do for engineers in their workflows? Most of us are familiar with the traditional approach of designing a 3D CAD model, and then we go through the tedious, laborious labor-intensive process of creating two-dimensional detailed drawings in which we then release. We do this to bring our product information into a human-readable format, which oftentimes leads to misinterpretation and rework. So with the MBD approach, 2D drawings are essentially eliminated. This is done by generating a 3D CAD model with MBD that puts all the relevant product information and full product definition into a machine-readable format, which can then be consumed by other tools in the digital workflow, such as aPriori. So MBD drives the model as the single source of truth and eliminates the drawing verse model conflict that often exists in manufacturing workflows.

Understanding the Digital Twin and How It Leverages 3D CAD Software Data

By now, we have heard of the digital twin or the digital thread, but what does that mean, or what does that look like? Because we’re moving towards digital workflows, it opens up a new realm of possibility and capability in leveraging machine-readable data. So it all starts with the 3D CAD model, and since we’re applying model-based definition and really adding those attributes of GD&T and PMI data, it allows us to branch off digitally into other workflows, other workflows such as CAM software. So this is probably the most common and most familiar with CAM software, we can read the CAD models directly, understand the tolerances and understand everything it takes to machine or manufacture this component. We’ve also heard of CMMs. This allows us to enable digitally where critical areas of the models are for inspection and testing purposes. Other tools like computer-aided engineering functions such as CFD analysis and FEA all leverage the 3D CAD model.

But we can think of this digital thread and expand in a different way, and that’s where aPriori comes into play. Here We can now use that digital model or 3D model to get accurate should costing analysis. We can also understand a full manufacturing process routing. With aPriori, we will also get provided with design-for-manufacturability insights or design-to-cost insights. We also now understand the CO2 impact on our sustainability initiatives and the emissions that are output by material and processes. But if we wanna take it to even a further level and we think about this digital thread, aPriori provides the groundwork to allow us beyond engineering to go what is known to a zero RFQ process. And we’ll cover that later on in the presentation as to what that looks like.

What we will discuss in this presentation is we’ll take a deeper look into how we can leverage the foundation that MBD provides to drive automation at different levels in our workflow with aPriori. So at level one, this serves as the foundation using the machine-readable tolerances and PMI data to reduce the number of required inputs and manual clicks that it takes to get to an aligned analysis. Level two is the acceleration of costing mechanical assemblies. And level three, aP Generate is the fully automated approach to read and write data into the CAD and PLM environment. And then, finally, level four, this is where we leverage MBD to support the zero RFQ process, which branches beyond engineering and into sourcing and procurement workflows.

So first, we’re gonna take a look at MBD, PMI, and CAD attributes. What is all of this? MBD is really the focus on applying what is known as GD&T. These are your geometric dimensioning and tolerances. These are the physical attributes of the model and the specifics as to how they need to be manufactured based on tolerances. PMI data is going to give you that secondary layer. So if you are doing secondary treatments such as anodizing or other posts inspection processes, you can find that in the PMI data as well as your materials, of course. And then there’s CAD attributes. So CAD attributes can live in many forms. On the aPriori side, we would use CAD attributes to help to drive the inputs of our processes, volumes, digital factories, all these can be mapped back and forth from our CAD system and into aPriori to help drive that automation.

What is CAD Mapping and How Does It Streamline Product Development Workflows?

So what does CAD mapping look like? So we’re gonna take a little greater deep dive into how do we map CAD attributes. And so in doing so, it’s a pretty simple process overall, but it does take a little bit of legwork in creating some custom properties on your CAD side. So what does that look like? If we take a look at this example here, we’re seeing a CAD parameters show in Creo. So what we need to do is create those particular parameters which are essentially going to drive the inputs on aPriori’s side. What we next do is we create a mapping file to map all this data from your CAD system and into aPriori. So what I’m gonna actually do next is jump into a demo and just demonstrate that flow of tolerances, PMI, and really the mapping of custom attributes to drive processes in aPriori. So let’s jump into the demo.

So here we are with a fully defined part inside of SolidWorks. And so, what I’ve created on the right-hand pane is a custom inputs tab. Now this is a capability strictly within SolidWorks. It could look different with other CAD platforms, but I thought SolidWorks does a good job of presenting how we can create custom CAD attributes if you will. And so what we have is a fully defined model-based definition component. We can see those particular tolerances and data planes assigned, but what I’m also going to input is the attributes that are going to drive the inputs for aPriori. And so, as we would here, we’ll select stock machining, we can indicate the digital factory we’re going to manufacture. I assign the material, and then next, we will apply our volume and batch size. And then any potential secondaries we may account for, in this case, I’ll anodize it. Once I have all this, I will go ahead and save it and then send it over into aPriori. And we can see here also on the left-hand side, we have all of our different GD&T and MBD attributes. As we go down the list, we can see what that tolerance status looks like on this model. And so, aPriori is going to read in these tolerances and apply the respective manufacturing processes. So let’s kick it over the fence to aPriori and then jump on into the software.

Okay, so here’s the component inside of aPriori, and if we take note on the left-hand side, we can see that we have the same inputs that we mapped over from SolidWorks. So we see our stock machining process group, our digital factory material and volume. And so, from here, I’m just going to go ahead and initiate the analysis, and it’s going to take a few seconds to calculate the should cost and DFM analysis insights that we’ll wait to complete here. And so now, with the analysis complete, we can begin to look at how aPriori has now brought in all those tolerances that were specified on the model. So if we go over here under the design guidance tab, we can see here in our tolerances and finishing operations that aPriori has collected those same tolerances that were defined inside of SolidWorks. So we can see here as we go through, we found all the different tolerances that were defined in aPriori as applies those as it pertains to the manufacturing process. So we can also note that it’s important to understand not every tolerance is going to drive a specialized finished operation. What it may do is slow down how the the machining rates occur or how fast we’re cutting or removing material. However, certain tolerances may push us beyond a threshold in which we’re driving specialized finished operations.

So we can see here aPriori is also going to indicate which tolerances are doing such. So we can note here that we have this particular tolerance on this bore. And so, aPriori is going to offer up the initial tolerance as well as the basic machining limit to allow the engineer to make decisions on how those tolerances are impacting manufacturing downstream. The same can be said for this specific tolerance that we’re driving a grinding operation due to the circularity that was called out on that specific feature. And again, aPriori will indicate the current tolerance value as well as that basic machining threshold. So the nice, the power behind MBD is all this was essentially automated versus the manual option of applying the tolerances through the interface. Still a capability with aPriori, but really just increases the amount of time it takes to get to that accurate analysis. So with model-based definition, we are now able to really reduce the number of clicks that are required to get to a calculated should cost analysis, and really understand how we can now leverage this and bring it to the next tier through automation. And so we’ll look at that later on with aP Generate.

Using aPriori’s Mechanical Assembly Functions and Critical Product Data

The next level of automation that we’re gonna take a look at is mechanical assembly. So in today’s aPriori by default setup, what we identify is simply pick and place. And we know that manually we need to apply other things like our hardware, things like rivets, nut plates, screws, nuts and bolts. But there’s a better way to do that. And so what I want to take you through is how we are going to leverage particular CAD attributes to drive the automation on aPriori’s side to be able to reduce the number of clicks we need to make to identify these assembly processes. So what does that look like? So I’m gonna share some attribute mapping logic, and it’s very simple, aPriori is going to search for the component numbers, and then it’s going to identify that these components reflect particular hardware attributes. And so then we can now apply the mechanical assembly functions for each of these pieces of hardware.

So what does that look like in the backend of aPriori? Well, it’s pretty simple. We create lookup tables, and in these lookup tables, these are going to serve as the math ways to the part numbers that we are searching for. So, in this case, on the window here, we could see in the configuration of our digital factory that we’ve configured a particular set of component numbers, which are going to identify different hardware features, things like rivets, cotter pins, nut plates, and bolts. At which case, once those part numbers are pulled in through aPriori, we are going to automatically apply the mechanical assembly process. So it’s a very simple thing to set up. There will be some legwork in pulling out all those part numbers and identifying, but in that initial legwork, it’s really going to cut down the time in costing and mechanical assembly and really just drive forward to focus on other areas.

Next, we have the complete automation solution with aP Generate. So what is aP Generate? Well, it all starts with the design engineer. Design engineer lives in their CAD and PLM environment. And through model-based definition, we’ve now built the confidence we need to automate this process. To do so, we use the software functionality within cost and site connect to build and customize our workflows. Within these workflows, we define which attributes get mapped back into the PLM and also how we handle our post-processing, which often comes in the form of an automated email notification. Within this email, a design engineer will receive a list of the components that they’re working on, highlighting any outliers that may exist. These outliers may come in the form of a high-risk DFM component or perhaps a part that’s above the target cost. From there, the engineer can then click on hyperlinks to access aP Design. Within aP Design, they can now investigate those insights and begin to make design decisions to figure out whether or not there needs to be any changes to the CAD model.

So what is automation and focus? So what we’re really going to focus on here is what does it take to get to automation? And so it’s pretty simple. It starts with a 3D CAD model. And in that 3D CAD model, we’ve built this foundation through confidence of our GD&T and PMI, and CAD attributes, aPriori needs to identify what inputs are required to initiate that analysis, and those are our basic inputs like our digital factories, our process group, our volumes, and batch size. And those are really the key inputs we’ll need to drive that analysis. From there, aPriori’s output is we get to choose which variables we then want to map back into our PLM. And so now we are able to leverage particular data points in our PLM, and who knows where you may go from there, you can actually push back that data into your CAD models, and I’ll be doing a demonstration here shortly to show exactly how we do that.

So in the ability to push data back into our PLM system and into our CAD models, we’re really able to expand our digital footprint. So with today’s CAD models, we know we have, again, that GD&T, the PMI data in terms of our materials. We can even set datum targets for inspection requirements. So we’ll apply things, other additional information such as anodizing or surface finishes in general. So with aPriori, we’ve now grown this digital footprint. This digital footprint now comes with a should cost. And so now we have the ability early on in the design process to understand are we meeting our target goals, or perhaps do we need to redesign this based on a different process group that we want to analyze? And we have confidence in these numbers we’re receiving because the model-based definition is really hitting on everything that is required in terms of manufacturing this component. We can also look at targets such as mass. In aerospace, mass plays a big role, and in understanding if we’re at our target, we can then use aPriori to analyze further different, explore different materials or different processes that reduce the weight of the component.

We can also get actionable design for manufacturability ratings. Early on in the process, we want to drive out any issues that are gonna cause issue downstream. And so, with aPriori, we can quickly engage and understand which components have a higher risk or are more prone to manufacturing errors or interpretation. And so aPriori will help you to drive that in terms of your best practices.

And then next is really the impact on CO2. This is a new initiative for a lot of manufacturers and companies however, there hasn’t really been too many tools available in which we’re really truly understanding our CO2 emissions impact. And so aPriori has that ability to calculate that and allow us to make those early decisions that we know will make for a more sustainable future.

Exploring the Power of aPriori Cost Insight Connect for PLM Manufacturing Data

So now what we’re gonna do is we’re going to jump into this level three demo of aP Generate, and we’ll focus here on the connection tool, on Cost Insight Connect, and understand how the data gets mapped back into our PLM system. So let’s jump into this demonstration. So starting with our Cost Insight Connect tool, this is really the administration tool in setting up a workflow that’s connected to the PLM environment. So the goal here is to really just show how do we build a workflow, how do we define the trigger, and how do we connect and map all the variables? So Cost Insight Connect is a a tool is where we’re going to achieve all of this. So let’s take a look, and what we’re seeing here is a collection of workflows that we have in our demonstration and testing sandbox environment. And so what I wanna do is just take you through the steps of what it takes to build out one of these workflows. So let’s go ahead and edit this particular workflow.

So the first step we need to take is obviously, name the type of workflow. You wanna maybe perhaps add a description, and then we need to choose the connector. So, in this case, we’ll be connecting to Team Center. The connector is really the definition of the PLM system. We also are going to use, in this case, a query definition process. So what I mean by that is we’re going to run this on a schedule. Below, we can identify that schedule, so in this case, we are running this every single day. Next, we need to use our query definition. So the definition entails how do we filter the components that we wanna run through an analysis? In this case, we use this for demo purposes, so we’ve identified that we want anything with this scenario name, Initial_CIG, will be automated on a daily basis. So as parts get checked in, if they have this scenario name inside the PLM, these are what will be triggered through the automation.

Next, we need to account for our costing inputs or inputs in general. These are the same inputs that we’re used to within aPriori that are required to run an analysis. So here we have to use our usual inputs like process group, material volume, back-size, digital factories, all those inputs are going to live inside of the PLM and then pushed into aPriori to drive the analysis. So we can see here we have a particular mapping rule, and in this mapping rule, we’re seeing that we have options where we can either run this as a constant, map it from a PLM environment, or have a default value if there is no value in the PLM.

Next, we want to identify the notification process. So we have a few options here where you can attach particular email summaries or summary reports that get attached. So this is where you’ll configure in a line the particular details that you wanna provide for that email notification. Finally, the publishing of results. So this is the writing of the data back into the PLM environment. And so we can choose which variables from aPriori we want to map back into the PLM system. And as usual, we can have these values push back in and choose and select which ones we want. So, in this case, we have our costing results, our rough mass, finish mass, labor time, piece part cost, investment, all of these, and some can all be written back into your PLM. We can also select which reports you wanna get attached as well. So if we wanna add our part cost reports to get attached into the PLM or our DTC summary report, we have the ability to do that. So this is all the steps it takes into building the workflow. On the back-end within PLM, it may be more involved in terms of you might wanna use this through particular defined workflows with your engineering department, but beyond this, it’s really a pretty easy mapping process.

So now that we’ve seen what it takes to build a workflow in Cost Insight Connect, here what I’m showing is some of the output results. In this case, this is the email notification. So with that same workflow, this is the email notification that gets pushed out. At the top, we can see we have a collection of part cost reports for all the components that were analyzed in that particular automation loop. Below, we’ll have a summary of all these components as well. Within that summary, live hyperlinks.

So these hyperlinks will jump us back into aP Design, and this will allow an engineer to quickly investigate any outliers, so for those that are at a high DFM risk in this particular case. So what we’re seeing here is this rollup of components, and we can see the particular, the component name, the process, the DFM risk rating, the aPriori should cost, and then the completion just to identify if there were any errors with the results. In this case, everything costed fine, and we can see here that we do have two outliers. So this is where if I was an engineer, I would want to begin to jump in and investigate further. Next, we’ll take note of the push and pull of the data that was set up in Cost Insight Connect to both drive the automation and also have the automation of writing back and publishing the data into our PLM environment. So right now, we’re looking inside of Team Center, and this is one of our sandbox environments that we use for testing purposes, so I apologize if it’s not the best looking or cleanest.

So what we’re looking at here is we’re on an item revision, and we could see here we’ve created an aPriori CIG tab. So in this tab is where we drive our inputs, so here we can see these were the inputs that were driving that automated analysis. We have our scenario name, which is again Initial_CIG, and that was the filter we used for this workflow. We also have our usual inputs like process group, VP or digital factory, material and volumes. So from here, this is what was queried for and then automated into aPriori. From there, after it ran through the analysis, aPriori then wrote the data back so we can see our outputs. And then, of course, we have a collection of components that were of the high, medium, or low scores. In this case, this particular component had a high DFM risk score, and we can also gain other intelligence, such as the fully burdened cost or cycle times and labor times associated with this particular component.

We can also see here, below is the attachment of reports that got attached and written back. Now we have a collection here because this particular component has been through numerous workflows if you will. But what I want us to jump into next is actually open up NX because we discussed earlier in the presentation about expanding the digital footprint, we can see now how we can push the data that was written back into aPriori back into our CAD systems. So let’s now open up NX and explore what I mean here.

Visual Reporting for Better Product Design Decisions

So here we are on NX, and this is an assembly in which all these components have been ran through the automation loop through aP Generate. And so, as we know, we’ve now written data back to each of these item revisions of these components. So this is a particular function within NX, and I think it’s actually very cool. But what I like to highlight is how we’re taking data from aPriori back into the PLM and now pushing it into the CAD model. And so, what could we do here with that type of information? And so NX, in particular, has a capability called visual reporting. So what we’ve done here is create a visual report which takes the data from the PLM and queries it searching for the risk rating of each of these components. So we can see here we have a report name called. So if I go ahead and activate this report, it’s actually going to indicate where my high-risk components are. So now, I’m completely living in my CAD environment utilizing the data that aPriori’s providing to be able to make more design decisions. So I, of course, can see that we have this one component that is a high risk.

So this, again, from a design engineer’s perspective, is where I wanna analyze my design. So just like I received in that email, I can open up this component inside aPriori and investigate what’s really going on from a design for manufacturability standpoint. And so the power of this is really great in terms of we can do other reports. I can run a similar report for target costs or even have another report for mass calculation in terms of what are not meeting particular thresholds that we set. So it’s a really powerful way to work, and it’s something you maybe never thought before, but it is really exciting. And I gotta credit my colleague, Jimmy Diedesch, for putting this report together. I cannot take credit for this, it was brain power, if you will.

How Zero RFQ Streamlines Sourcing Workflows in the Supply Chain

Next, we’re going to take a look at how zero RFQ can be implored through model-based definition and really how we build the foundation to go to a zero RFQ process. So what we’re looking at is a traditional sourcing process where we know it all starts with engineering, who has to release a technical data package, then provide that data package to sourcing. In which case, you’re going to launch their process of searching for suppliers to be able to manufacture. So this is a long-winded back-and-forth process, not only between engineering and sourcing, but sourcing and potential suppliers.

So once we have identified the project and the requirements, we are then going to send out for quote. It’s a long-winded process to get back those quotes and then eventually go to an agreement on purchase. So what can we do here with zero RFQ? Well, we know our common challenges, and that’s really the pressure to release data from an engineering perspective, the pressure to deliver data from sourcing perspective, and then the pressure to accurately interpret that data. With MBD and the model-based workflow, we can really move away from a lot of issues that we’re facing. Those issues are really the time savings on the engineering front. I speak from experience here when I know my biggest frustration as an engineer was detailing a CAD model, working in 3D, knowing that model was complete and I was happy with my design. But understanding I still had to undergo the long-winded effort of producing two-dimensional drawings in which that’s really where the pressure and bottleneck begins for engineers. The drawing process is long-winded, there’s a lot of back and forth, a lot of checking, and really what happens is, is you’re trying to interpret that single source of truth and bring it to an area that it could be human-readable.

With model-based definition, we’re now going to machine-readable. This enables sourcing and your suppliers to have confidence in what they’re quoting. In doing so, it’s going to relieve suppliers of that laborious interpretation of reading two-dimensional drawings and making sure they clearly understand the manufacturing process. With that, it’s going to give them confidence in their coding abilities. There’s a very fine line for suppliers in terms of margin of error. They can really take a big loss if they really didn’t interpret that work package correctly. It’s going to also help sourcing. Sourcing is going to be able to get out these drawing packages much faster. And because of model-based definition, all this has improved all these workflows across the board. So knowing that we’re improving the flow of data, what is the zero RFQ process? The zero RFQ process is the balance of unlocking the power of aPriori to digitally change the way sourcing and procurement work with their supply base.

At the heart of this capability, lives the 3D CAD model. The complete product definition enables aPriori to accurately align should cost and the manufacturing process. With the zero RFQ, both parties work to build and configure digital factories in aPriori, which are set to service the quotation and purchase order system. Confidence in the system is built as products are tuned in both parties essentially agree on margin, making this workflow win-win for sourcing and suppliers, dramatically reducing the back-and-forth communication of quoting and verification and minimizing any risk on a project. So what does that workflow funnel look like now? Well, we know that model-based definition is providing the foundation in terms of the accuracy and alignment, we have the ability to automate particular workflows. So MBD is really driving the data flow and communication in the sourcing process. With zero RFQ, we’re automating the PO process. And so now our workflow is much more simplified because we’re taking out multiple steps, we’re ensuring confidence through machine-readable data, and have both parties to agree that this is the most efficient and best path forward. So the ability to get to zero RFQ doesn’t happen overnight, and it really takes that foundation of being built, and we’re seeing how we may be able to implement that.

So to summarize where we stand with leveraging model-based definition and aPriori, I’m going to bring back up this pyramid and really just a reminder of how do we get to the pinnacle? And, of course, the pinnacle requires a strong foundation. That foundation is as we explored in this presentation, is leveraging model-based definition to really drive automation and take out the tedious process of manual interpretation. Through that, we can now accelerate our process at another level through mechanical assemblies. We now have confidence in our sub-components from individual component standpoint, but now we can take it to that next level and really accelerate the process of costing our assemblies. So in doing so, that feature that we shared earlier of mapping CAD attributes to drive hardware is going to help accelerate the costing of your mechanical assemblies. And if you really want to try to get to this pinnacle ultimately to zero RFQ, we can automate. So with aP Generate, we have the ability to automate everything we’re doing below. And now we know that in that automation, there’s confidence that because the foundation has been built, we are confident in what is getting automated. So it’s really going to limit any manual input or perhaps manual interpretation that is required.

And through that, if we wanna get to the pinnacle that, not everyone reaches the pinnacle, but if you wanna really get there, we were talking about zero RFQ. And zero RFQ, as we discussed this previously, is a substantial process in really speeding up, eliminating that tedious back-and-forth time-consuming process where both parties potentially lose on getting a bad deal or not satisfying the requirements of the contract. So now remember, the smarter the CAD model, the smarter the workflow. And hopefully, there has been many good takeaways from this presentation, but I believe the most important is focusing on building the foundation and moving toward model-based practices.

Thanks again for joining me in this discussion, I hope you’ve enjoyed your Manufacturing Insights Conference experience.