Systems Simulation: Far-off Future or Feasible Now?
May 8th 2013
Move over multi-physics. There's now a different field that qualifies as the bleeding edge of technology in simulation: system simulation. Many organizations are just starting to broach this field in an effort to better understand how systems perform holistically.
But what is it? Is it really feasible for practical use today? Let's dive in and discuss.
Setting Baselines: Systems Engineering
Terminology can be such a challenge. So before we go barreling off to talk about the questions above, let's set some baselines. To start, if you're unfamiliar with systems engineering, go check out its entry in wikipedia. It does a fairly good job explaining the key points.
Furthermore, if you want even more information on systems engineering, you'll want to head over to the site of the International Council on Systems Engineering, or INCOSE for short. They have developed a system engineering standard, both in terms of processes and definitions, that are fairly mature.
Between those two, you should get a good idea of what system engineering is all about.
Setting Baselines: System Simulations
OK, with that out of the way, we arrive at the next question: what is a system simulation? And, probably just as importantly, why is it different from the simulations being performed today?
The answer is actually relatively simple. Most simulations today address the engineering physics, such as statics, dynamics, thermodynamics and the like, that occur in relation to the mechanical properties and characteristics of a design.
Recent years have seen more technologies that apply to the electronic properties and characteristics of a design, specifically heat management related to fluid dynamics and interference via electro-magnetic fields. However, more simulations are emerging around the behaviors of processors and printed circuit boards, more along the lines of timing.
Of course, there have been simulations and test that would track the behavior of software applications and systems for years. Funny enough, these tools have been in place at software developers for quite some time.
So what's missing? Well, putting it all together. No. Seriously. Getting these three disparate kinds of simulations to work together in no small feat because the result of one simulation affects the outcome of the other ones, especially when software feedback loops are applicable. And today's most complicated systems really heavily on software.
Signs of Progress
So, are we almost there yet? Well, let's discuss progress so far and maybe you can be the judge. There are two main fronts of progress to consider.
Model-Based System Engineering Definitions
The first front to assess is the definition of standards. Wait! I know. Standards bodies haven't always been known to move at lightning speed. I'll give you that one. However, promising strides have been made here. Specifically, NAFEMS and INCOSE announced a collaboration (Press Release announcement here) that creates a new working group called the Systems Modeling & Simulation Working Group (SMSWG). You can download a presentation that roughs out the vision behind this working group here. Below is a glimpse, however, of the vision of how such simulations are composed.
Simulation Software Capabilities
The other front that holds some promise is the capabilities being developed by software providers. There are a few notable examples I'll highlight here.
- Dassault Systèmes' Dymola uses the open Modelica modeling language to capture system behaviors. It can work in CATIA or independently and offers a wide range of capabilities. Given DS's 3DEXPERIENCE vision, I'd be very surprised if they didn't come out with some specific system simulation tools that more completely support the process soon.
- CyDesign is the new kid on the block that uses 2D and even 1D modeling to crank out hundreds and even thousands of system simulations that is available in the cloud. They're one to watch.
- Wolfram is making a run in this space as well with their SystemModeler. They use a drag and drop functional interface to build out mathematical models behind the scenes.
Now, there are more software providers working in this space. Much more. Keep your eye out and you'll see more progress on this front. The key thing to keep in mind is that there are solutions for this field emerging from everywhere, including traditional PLM players, startups and software providers that have never serviced this industry before.
There's a problem to be solved. It looks like software providers are lining up to take a crack at solving them.
Summary and Conclusion
Here are my thoughts.
- System simulations capture the composite behavior of systems composed of mechanical, electronic and software components. Each disciplines has simulation tools. To date, they really haven't been connected together.
- NAFEMS and INCOSE have partnered to provide some definition around what a solution for system simulations should look like. Investigate it. Get involved in the working group. That is an important front to track.
- New software tools are emerging that have impressive capabilities from various providers. They're not perfect, but they are progressing.
Are we there yet? My answer is no. But we're not decades away either. Solutions are nearby. A little more progress and they'll truly be feasible.
What do you think? Any particular areas that need more focus than others? Any other software providers that deserve notable mentions for system simulations? Sound off. Be heard.
Take care. Talk soon. Thanks for reading.