Skip to Main Content
Talk Beginner

Ecosystem of Opensource tools/libraries/softwares in cyberphysical product design and development

Proposal status is Rejected
Session Description

When we say opensource software, we immediately picture Linux, kubernetes, postgres etc. The tools that are widely used to build cyberphysical products, which are known as CAE (Computer Aided Engineering) tools do not get mentioned at all in the FOSS conferences. Some examples of the cyberphysical product simulations are a. jet engine performance testing, performance of chip cooling inside laptop, crash-testing for a car, or blood flow simulation through a stent. There is a plethora of tools available in CAE and one must choose the right one depending on the physics being simulated, application being designed etc.

In this talk, I want to introduce the open source projects and tools in the CAE arena. I'll show how the ecosystem is organized: OpenFOAM as the generalist workhorse. Around it are bunch of specialists, each purpose-built for physics/use-case: UCNS3D for high-order, shock-dominated compressible flow; Basilisk for droplets and interfaces; PyFR for GPU-native, scale-resolving turbulence. The structural world mirrors this: Code_Aster as the deep generalist for nonlinear mechanics, fatigue, and fracture; CalculiX for thermomechanical analysis; Elmer for multiphysics.

In my second half, I will zoom out. I will present how simulations as an industry is a multi billion dollar software market (I will include more details with breakdown in the presentation) and that is largely dominated by handful of players like Ansys (now part of Synopsys), Siemens, Dassault Systèmes, Altair (now part of Siemens). On paper, open source could serve a large share of this market. In practice, it does not. Having spent more than 7 years working in the industry I can say that this is mainly because of gap in adoption. Commercial vendors sell support along with software. A dedicated team that can picks up the phone when things go wrong.

So I'll close with a proposal, and it has two halves.

  1. Working way of contribution to OSS: I would propose a way of working where the software engineers team up with domain experts, where the software folks bring the software expertise and the domain experts bring the physics and the real industrial requirements, and together they close the gap between what industry needs and what the tools deliver today.
  2. A support framework: an ecosystem where companies adopting open solvers have somewhere to turn when there is a need for a dedicated support. One way to also solve this in the long term is to train the young engineers with open source tools so that the market has a pool of engineers who already has the software knowledge and expertise of OSS tools/products.

Once we have both the above pieces, open-source CAE becomes a genuine, trustworthy foundation for cyber-physical product development.

Key Takeaways
  1. CAE is an ecosystem, not one tool. Depending upon the physics to be simulated, one needs to choose a specific set of tools.
  2. It is a multi-billion dollar (conservative) market and there is consolidation happening all the time
  3. Open solvers stay underused because there's no one to call when they break.

References

Session Categories

Introducing a FOSS project or a new version of a popular project
Technology architecture

Which track are you applying for?

Main track

Speakers

Aniket Kulkarni
Aniket Kulkarni Founder | Curlscape Solutions

Aniket Kulkarni is an experienced AI leader, entrepreneur, and technologist with expertise in AI; specifically AI Agents using OSS, open source small language models. Aniket is passionate about exploring technology for creating efficient, scalable, and accessible AI solutions that positively impact communities.
Beyond his professional pursuits, Aniket is dedicated to promoting knowledge sharing within the AI community, driving impactful conversations, and mentoring future leaders in technology