An Inside Look at Full-Vehicle EMC Analysis
Automotive platforms are becoming increasingly complex, with thousands of cables, densely packed electronics, and growing electromagnetic compatibility (EMC) challenges. To understand how engineers are tackling these enormous simulations, Electro Magnetic Applications, Inc. (EMA) Co-CEO Tim McDonald sat down with Sebastian Soldwisch, EMA Senior Scientist and Product Manager, to discuss how modern simulation technology is enabling true-platform vehicle analysis.
Read the full interview below.
Meet the Scientist Expanding What’s Possible in Simulation
Tim McDonald: Sebastian, tell me about what you do at EMA?
Sebastian Soldwisch: I’m a Senior Scientist and Product Manager. That means a lot of my time is dedicated to helping develop new workflows and leading new physics solver initiatives to progress our software products, Ansys Charge Plus, Ansys EMC Plus, and STK Shield Plus. Overall, just trying to tackle new frontiers of what we work on in the electromagnetic space.
Besides that, I help direct the solver and GUI development by outlining new features for tools and just making sure that everything meets our expectations on quality and performance.
TM: You do a lot across many domains. What’s your favorite part of your job?
SS: My favorite part of the job is definitely figuring out new ways to simulate something we’ve never done before. One of the beauties of our tools is that we’re solving Maxwell’s equations, electromagnetics on a fullwave level. That means the possibilities are endless on what we can simulate. Finding out new ways to solve new problems is just very rewarding.
Beyond the Shell: Modeling the Entire Vehicle
TM: How have you helped customers that have automobiles with a lot of cables in them specifically?
SS: The unique thing about EMC Plus is the combination of our FDTD solver and transmission line solver. That means we can tackle full platforms in a very computationally efficient way. Just being able to assign cross sections of cables to mesh routings means that we can handle thousands of cables in a single simulation.
Computational efficiency from hybridizing that transmission line solver and FDTD solver are even more accelerated by our GPU capabilities. That just means we’re solving these gigantic problems in timescales that are minutes to hours, rather than days to weeks.
TM: You’re talking about full platform, I hear people talk about full platform, however when you look at what they did, it’s a very simplified shell of a vehicle, a lot of the structures in the CAD aren’t there, there’s just a couple of cables. Is that kind of simple shell what you mean when you talk about full platforms?
SS: Definity not. Another advantage of how we approach this is that the FDTD meshing is very forgiving. That means a minimal amount of calculating is needed to include that full mechanical CAD, not just the shell. All the chassis, the seats, the dash in their ECU enclosures have very accurate harness routing throughout each subsystem. So, when we say full platform, it’s true full platform.
TM: You said thousands of cables earlier. That sounds like a stretch. I mean, nobody really simulates thousands of cables, do they?
SS: We’ve been doing thousands of cables for years now, maybe it’s more accurate to say a decade or more. With our lessons learned from doing thousands of cables on aerospace platforms, we’ve been applying that approach to automotive.
That hybridization of the transmission line and FDTD means that we can have these complex cable bundles, literally dozens, sometimes even pushing 100 per a single bundle. When you extrapolate that to how the harness is routing throughout the entire vehicle, that easily reaches thousands of conductors, shields, and so on.
TM: If you put that kind of complexity in, I imagine you’re approaching a billion cells. How could that be computationally efficient?
SS: It is sometimes approaching a billion cells. But with GPU acceleration, it’s computationally efficient to handle that. Anecdotally, some of the full platform automotive models that I’ve been doing recently were a half a billion cells. But we’re still running radiated immunity in a matter of an hour when doing multi-GPU acceleration. When we’re getting on the scale of a billion cells, multi-GPU helps us keep that in the well under a day range.
How Agentic AI is Transforming Electromagnetic Simulation Workflows
TM: I know a lot of people I work with are switching to more agentic workflows where they’re using Claude or Codex with what they’re doing. Does that touch any of the workflows that you’re talking about?
SS: 100%. Right now, the way I’m mostly using different AI tools like Claude is for post-processing. The huge initiative at EMA is being able to use that more in that kind of agentic fashion, of helping with pre-processing, setting up the simulation, actually queuing simulations for a sensitivity analysis, maybe an optimization and then getting into that post-processing. It’s here, it’s now.
A Practical Path to Full-Platform Simulation
TM: For a lot of people this is going to be new because I think full platform simulation hasn’t made it everywhere yet. I think it’s just getting started in the world. How would you recommend if you’re an engineer at a vehicle company responsible for the full platform, how would you get started with simulation and then flow from that, and how you start adding some of the agentic approaches?
SS: I think the first thing is that you need to figure out what your end goal is. If you’re brand new to simulation, a lot of people think the first step should be ‘I need to match exactly what I’m getting in measurement versus my simulation.’
When we’re talking about the full platform, just because we know that we’re the best at full platform, doesn’t mean it’s not complicated. It’s inherently complicated. But if you figure out your goal and you think of milestones like, what engineering insights I can get as I build up to my full fidelity of a full platform with thousands of cables, every single bit of MCAD in there, then you can extract engineering insight through starting with simple models.
Let’s say maybe you are starting with that shell with full bundles in it, but that’s something you can start, simulate, and set up in less than a day. Even for some beginners, if you do your first project, as you build up complexity, extracting engineering insight as you build that up, then you have a foundation of how to use our products that can then lead you into best paths forward for actually knowing what you want to do with those agentic codes. I think having that fundamental understanding of it is important when you’re trying to go for that advanced simulation.
The Future of Full-Vehicle EMC Starts Now
As vehicle platforms grow more complex, full-vehicle EMC analysis is becoming essential for reducing risk and accelerating development. What once seemed impossible, simulating an entire vehicle with full CAD, thousands of cables, and hundreds of millions of cells, is now achievable thanks to hybrid solver technology, GPU acceleration, and emerging AI-driven workflows. As a member of the NVIDIA Inception program, EMA is helping advance the next generation of high-performance simulation, giving engineering the computational power needed to analyze increasingly sophisticated vehicle platforms with unprecedented speed and scale.
As Sebastian explains, success comes from building a strong simulation foundation, gaining engineering insight at every stage, and gradually increasing model fidelity. Whether you’re just getting started or pushing the limits of large-scale EMC simulation, the tools, technology, and expertise are available today and EMA is ready to help you every step of the way.
Ready to Get Started?
EMA’s team of application engineers and scientists can help you accelerate your full-vehicle simulation journey. Contact EMA now if you’re ready to:
- Discuss your vehicle EMA challenges and simulation goals
- Explore EMC Plus and learn how hybrid solver technology enables true full-platform analysis
- Work with EMA experts to build a simulation workflow tailored to your vehicle and certification requirement
- Evaluate GPU-accelerated simulation strategies to reduce turnaround times from weeks to hours
- Discover emerging AI-driven workflows that can streamline model setup, optimization, and post-processing.
Now is the time to learn how full-vehicle EMC simulation can help you identify issues earlier, reduce development risk, and bring products to market with greater confidence.
