The evolution of AI is affecting every corner of society and industry, and that includes simulation. Engineers have long taken advantage of machine learning techniques in their simulations, but recent advances in hardware and technology have unlocked new potential for innovation, powered by AI and ML.
Christina Feist, the Generative Experiences Role Portfolio Director in SIMULIA R&D, is using AI and ML in her work to support the simulation user base and translate that into AI tools that make their work easier.
Faster Design Cycles with the Power of AI
Every time a design is iterated, it needs to be analyzed again in order to understand the effect of design changes. This re-analysis time can be a limiting factor in the speed of the design process and the total number of iterations possible.
SIMULIA is delivering Virtual Twin Physics Behavior models to accelerate development and enable faster analysis. They are based on surrogate models trained using machine learning on simulation data and can “predict behavior in nearly real time,” says Feist. “The time required to evaluate design alternatives is dramatically reduced, and teams can explore more concepts, iterate with greater confidence, and converge on higher performing designs – much faster than with traditional simulation alone.”
Making AI Intrinsic to the Simulation Workflow
Rather than having disconnected AI tools and siloed data stores, the SIMULIA approach integrates machine learning into the simulation process itself. “We no longer need to simplify the physics to make simulation approachable,” explains Feist. “Instead, AI enables a more intuitive and guided experience that remains rooted in high-fidelity engineering.”
The Virtual Twin Physics Behavior runs on the 3DEXPERIENCE platform and is built on data from mature, proven SIMULIA technology. This tool links simulation and modeling (MODSIM). “The result is a more inclusive simulation environment that’s accessible to a broader range of users while also supporting sophisticated workflows.”
Encapsulating Knowledge and Know-how
One of the biggest challenges when implementing simulation in a development team is democratization: ensuring that everyone who needs simulation has access to it. For a successful simulation, the solver and simulation environment must be set up with the correct settings, and this usually requires expertise on the part of the user. Feist adds that AI “lowers the barrier that has traditionally made simulation difficult to deploy outside dedicated simulation departments.”
With Virtual Twin Physics Behavior, simulation know-how is encapsulated inside a model trained by an expert simulation analyst. “The methodologies become intrinsic to the workflow rather than something that the user must learn or manually reproduce. Users can expect a more comprehensive, collaborative, performance-driven workflow,” says Feist. “Designers have greater confidence and analysts amplify their expertise across more design cycles.
Secure AI Protects Sensitive IP
As users become more aware of how AI tools work, the issue of intellectual property has become increasingly pressing. Confidential data should not be leaked by the training process, nor should the model’s outputs infringe on protected IP from another company. SIMULIA takes a unique approach to managing customer data within AI. “Our approach is to treat them just like we would traditional simulation data – as the customer’s intellectual property,” explains Feist.
“These models are trained in a secure environment with full 3DEXPERIENCE platform traceability using only customer-selected data sets with no cross-learning across organizations.” This means that the data is governed and fully auditable on the 3DEXPERIENCE platform, allowing users to be sure of their data sources. “Customers can realize the benefits of AI and machine learning while maintaining confidentiality, integrity, and the controlled use of their engineering data.”
Key Takeaways
- Virtual Twin Physics Behavior uses the power of AI to build machine learning models trained on simulation data from the SIMULIA simulation portfolio.
- Virtual Twin Physics Behavior is built on the 3DEXPERIENCE platform, allowing users to take advantage of the power and accuracy of industry-leading physics simulation tools in the SIMULIA portfolio.
- Virtual Twin Physics Behavior protects user’s IP and uses the unique traceability provided by the 3DEXPERIENCE platform to ensure that data is fully audited and not infringing.
- It will bring big benefits for simulation users, increasing the speed of analysis and accelerating time-to-market.
- It also promises to help designers and engineers who want to use simulation in their daily work, by encapsulating expert know-how and breaking down the silos between modeling and simulation (MODSIM).
Interested in the latest in simulation? Looking for advice and best practices? Want to discuss simulation with fellow users and Dassault Systèmes experts? The SIMULIA Community is the place to find the latest resources for SIMULIA software and to collaborate with other users. The key that unlocks the door of innovative thinking and knowledge building, the SIMULIA Community provides you with the tools you need to expand your knowledge, whenever and wherever.
Katie is the Editor of the SIMULIA blog, an expert in engineering and scientific communication, and a strategic digital marketer. Katie has a BA in English and Writing from the University of Rhode Island and a MS in Technical Communication from Northeastern University. She is also a proud SIMULIA advocate, passionate about democratizing simulation for all audiences. Katie is a native Rhode Islander who lives just outside of her favorite city, Providence. She is a true New Englander and loves sharing all the cool things NE has to offer. She enjoys a variety of hobbies including history, astronomy, science/technology, science fiction, true crime, fashion and anything associated with nature and the outdoors. She is also mom to a 6-year old budding engineer and two crazy rescue pups.