Tech4PD: Episode 2
Simulation driven design has been proven to have great value for the end stage of product design. Many companies are now moving simulation early on in the design process to reduce rework so changes can drive design. Knowledge of engineering physics, simulation methods, and CAD software are all required for success. Should engineers absorb this work or should analysts take it on? Watch this episode and learn about who could close the simulation gap with the smallest learning curve.
The topic of this episode is an interesting example of a product development silo collapsing. Historically companies could afford to have one group work exclusively on conceptual design, another on detailed design, and yet another on validation and verification. Due to many market pressures namely reducing product development costs, minimizing time to new product introduction, and increasing product quality, the blending of these groups and responsibilities is becoming more commonplace.
At PTC, we think that this trend will only continue and as a reactionary step we need to enable it. We believe that analysts should be more versed in conceptual and detailed design tools. Likewise, designers need to be better versed in analysis tools. We are also sympathetic to workload concerns. It is not reasonable to think that a designer is going to have time to understand and use yet another software tool, especially one that is unaligned to their current work tasks. As a result these tools need to be seamlessly integrated into the design process and easy to use by analysts, engineers, and designers.
To solve this problem, PTC is focusing on an application strategy, whereby any role (analyst, engineer, designer, etc.) can view the design data using their own lens. This can be seen in the application strategy behind PTC Creo, specifically with the PTC Creo Behavioral Modeling Extension, and with the simulation technologies that integrate with PTC Creo, like PTC Mathcad, PTC Windchill, and PTC Creo Illustrate.
We are also enabling easy switching between these applications so the user can get different views quickly. For example, a designer should be able to take their part or assembly under development, and evaluate how it performs under a load or with a motion limitation. This is where modeling and simulation technologies bear fruit. These technologies enable the design process and provide quick feedback on design performance. This also enables more detailed design analysis down the line as the parts and assemblies are refined. As product complexity grows these types of design and analysis integration improvements must happen. You simply cannot create complex products without knowing the complete context of the design.
Jim: Welcome to Tech4PD, the technology show for product development. I'm Jim Brown. I'm one of your hosts.
Chad: And I'm Chad Jackson. I'm your other host. Today we're going to be talking about an old but very good topic, simulation-driven design.
Jim: Great. Great, so here's how this show's going to work. First, we're going to start off and we’re going to share some background and set the stage. Then what we'll do is we’re going ahead and debate our views on this particular topic. Chad and I often have different opinions. We'll be sharing those with you. Then what we’re going to do is we'll take a look into the crystal ball and give you our idea of what we see happening in the future in simulation. Finally, we'll take a look at the consequences from last episode's debate.
Chad: That's right. Actually for this episode, the consequence, and you'll vote on who the winner is, the loser will have to dress up in a Star Trek uniform, complete with Vulcan ears, and go to a conference or a business meeting.
Jim: We're looking forward to see Chad do that. Anyway, let's get started. To set the stage with simulation-driven design, what we're seeing people try to do now is really try and move simulation earlier into the design process. Not just at the end for validation and verification, a sort of final check, but earlier into the process so we can still make a lot of changes and use it to drive design, as opposed to putting it on the back end where all of the windows of opportunity have closed because of decisions that have been made and locked in.
Chad: It's certainly not easy.
Chad: In my mind there are four main things you need to know if you're going to be performing simulation-driven design. One is engineering physics. The second is the simulation method being used. The third is how to use the simulation software, and the fourth is how to use the CAD software.
Jim: Absolutely. We've seen a lot of changes and some evolution over the last four or five years where we've seen some blending. We've seen embedded analysis tools within the software and not just on the same menu bar, but actually in the same environment so it's much more familiar. We've seen some more hand-holding, some guidance, some wizards, that really are trying to make it easier for the designers to really approach this process. They may not get it perfect, right? Even if it's just directional, that's a lot better than waiting until it's too late to change, right?
Chad: Absolutely. Some of the guys who have been making those changes, you look at people like MSC, ANSYS, ESI Group, Altair Engineering. They're taking their simulation tools and actually putting them in the CAD environments.
Jim: Right, and the bigger vendors as well, not that they're all small vendors. The larger Suite providers in the design tool space, you know the PTCs, Dassault Systems, Siemens PLM, and Autodesk have also moved tools into their own solutions. So we've see this blending going on.
Chad. Right. The impact of that has been pretty important. When you start to see simulation capabilities being integrated into the CAD tool, it really starts to address the last two challenges. How do you learn how to use the simulation software and the CAD software? The guidance that you're talking about really starts to address the first two challenges around what is the engineering physics that you need to know about to perform the simulation as well as the simulation method? There's been a lot of progress on these fronts.
Jim: Absolutely. I think that sets the stage. Let's have a little debate. To kick things off, given today's capabilities, there's two paths to go. One is to take your designers and move them closer to your simulation. The other is take your analysts and move them closer to design. Both valid paths.
Chad: For me it makes a ton of sense to take simulation analysts and get them more involved up front in the design cycle. Of course, the issue you're going to run into there is that they have a huge queue in terms of work for verification and validation. How do you address that? I think if you can automate what they do, make them more efficient in that realm, then you can move them more up front. For me, it would be great to have engineers involved, but considering those four challenges, I just think it's too much.
Jim: I think it's too much, too, but I think it's too much to ask our analysts to pick up and do more. Yes, you should make them more efficient. There needs to be more capacity. You don't want to add hand-offs. If you want to drive efficiency, you don't want to add more hand-offs back and forth between engineers and analysts so the design floor capability to me is really much more important. Even if it just means it's directional analysis, it may not be as accurate, but it doesn't need to be either. So I still think that's the way to go. There are learning curves either way, right? Analysts aren't exactly usually as versed on a CAD tool.
Chad: That's been a big concern traditionally, but if you look at what's happening in the market right now, there are a lot of new tools that have come out to help simulation analysts prepare geometry for simulation. You look at what, for example, SpaceClaim is doing with direct modeling, as well as some other ones, some guys like PTC and Siemens as well, making huge strides there.
Jim: No doubt, but at the same time I think they're also huge strides that make it easier for designers to be doing simulation. Let's just make it clear here. If you're running an engineering department right now, what would you do? Would you take your analysts and move them closer to design, or would you take your engineers and enable them with simulation?
Chad: For me, I would take simulation analysts and move them more up front. They're a proven commodity. I know they can make an impact. For me, it's an easy decision.
Jim: It's an easy decision for me, too. Proven commodity, but also a precious commodity. There aren't enough analysts available to do that. We've got to get more capacity from the designers to get it done earlier without the hand-off.
Chad: Luckily, they get to decide.
Jim: But they know I'm right.
Jim: We'll see. Thanks.
Chad: All right, now let's shift gears. Jim and I are going to peer into our crystal ball and predict a little bit of the future of simulation-driven design in the industry. Jim, you go first.
Jim: First thing that I'm seeing is definitely cloud computing, just huge amounts of resource that can be made available today on demand. The ability to access lots of processors, lots of memory, in parallel, at the point that you need it without having all of that sitting around, I think is just really compelling.
Chad: Yeah, and certainly simulation has been computationally intensive. It's something that could have a big impact. An area that I want to talk about that usually you might not associate with the simulation is mobility. You think a lot about compute resources, but with mobility you could actively monitor, watch the progress of using compute resources like the cloud. Also, it's a good way to collaborate with others so you don't have to be sitting in front of your desk to share some results or the progress of your simulation. I think that's going to have a big impact, too.
Jim: Speaking of visualization and sharing, I think another thing we're going to see definitely is data management. With so many more simulations going on, particularly if you're using the cloud and sending multiple data sets across with some different design options, bringing them back so you can make some trade-offs. Being able to manage all of those, and hopefully tie those back to the rest of the engineering process, I think is going to be a really interesting thing. We're starting to see a lot of that happen along with automation and really trying to make each of those steps faster and more efficient, as you were talking about earlier.
Chad: It's going to be great when all the simulation analysts can move forward in the design cycle.
Jim: Even more importantly, this is the fun part of the show. We're going to find out what you, the audience, voted in terms of last week's debate, and we're going to have you take a quick look now at the consequences.
Chad: Let's see.
Jim: Now is the time for the consequence phase, and by the fact that you're seeing my lovely face here instead of Chad, that means that you decided to vote for granularity as opposed to integration.
[sound of electric toothbrush and laughter]
Chad: I still believe in integration. Thanks for tuning in today. We'd also like to take the opportunity to thank our sponsors, PTC and Mentor Graphics, for backing the show.