By Shawn Wasserman
The first, of likely many, AI-powered simulation tools to come in 2024!
Disclaimer: Shawn Wasserman is a former employee of Ansys and holds minor stocks in the organization.
Engineering.com predicted that 2024 will bring an AI revolution within the engineering simulation software space. Though announced in early 2023, Ansys SimAI is the first to solidify that forecast, as it is now officially added to the simulation leader’s portfolio.
Ansys SimAI is for engineers and laypeople alike, says Ansys. It promises to offer physics-agnostic simulations via machine learning and a software-as-a-service application. The aim is to merge the predictive performance of Ansys’ simulation technology with fast and accurate generative AI.
A preview of the simplified Ansys SimAI user interface. (Image: Ansys.)
As products become more complex and development times shrink, the demand for engineering simulation software that can accurately and quickly return results is increasing. By training generative AI and integrating it within the design process, users can get simulation results faster to reduce development time and to enable the testing of more of the design space.
“Our customers across industries are becoming more reliant on generative AI to improve their design processes,” said Shane Emswiler, senior vice president of products at Ansys. “The increased demand for cloud-native solutions signals momentum in the shift from siloed workflows to a more open, collaborative approach to simulation. Ansys SimAI, in combination with other Ansys solutions, opens a world of possibilities — helping organizations develop comprehensive, end-to-end processes for diverse applications with inherent time and cost-saving benefits.”
This might sound similar to Ansys AI+, which was also announced late last year. The major difference is that Ansys AI+ will extend and add AI and machine learning features to other tools in the Ansys portfolio, starting with Ansys Granta MI AI+ and Ansys optiSLang AI+. Meanwhile, Ansys SimAI’s physics-agnostic nature will mean that it can be trained on various types of simulation data spanning the needs of numerous industries.
“AI+ are modules that extend the capability of our existing products by leveraging AI/ML technology,” confirms Ilya Tolchinsky, Lead Product Manager at Ansys, to engineering.com. “They cannot function without underlying products like Ansys optiSLang.
“Shrinking development times and increased complexity also means that more people specializing in various disciplines can benefit from simulation technology. To address this, Ansys SimAI offers a simplified user interface that is designed for users without coding or deep learning experience. As for simplifying the typical simulation process for non-experts, Ansys SimAI only cares about the shape of the objects being simulated—not the typical geometric parameters used to define a design. This also means that the tool can predict a larger design space even if there are inconsistent shapes in the training and input data.
“With Ansys SimAI, we will be able to easily test a design within minutes and rapidly analyze the results, ultimately redefining our digital engineering workflow and reshaping our perception of what is possible,” said William Becamel, expert leader in numerical modeling and simulation, Renault Group. “By enhancing simulation speed, we can explore more technical possibilities during the upstream phase of our projects and reduce the overall time-to-market.”
Users train Ansys SimAI using previous data generated by Ansys and other simulation tools. The data is hosted on a cloud infrastructure that is designed to keep intellectual property secure and private. Ansys reports that SimAI features an open ecosystem, intuitive interfaces, and the ability to predict product performance in minutes. The company hopes this will lead to more people using simulation technology who would typically find it too esoteric.
Filed Under: AI Engineering Collective, Commentaries • insights • Technical thinking, AI • machine learning