The improvement of testing tools, modeling and agents based on artificial intelligence are bringing the industrial sector closer to the center of the debate on the possible applications of technology for efficiency gains and cost reduction. The applications, called physical AI, are already found in projects in the automotive and aerospace industries.
“In my opinion, physical AI is the main application of AI in engineering available in the world today”, said Siemens senior vice president, Yeshwant Mummaneni during the Realize LIVE event held in Detroit, United States, this Monday (1), which the report from InfoMoney accompanied. “It is literally turning weeks and hours of workload into seconds.”
Also read:
Study abroad
Upgrade your career!
Physical AI uses the massive data loads of artificial intelligence models to enable systems to understand, reason, and take actions based on the behavior of objects in the physical world. The idea is that programs are able to predict the behavior of products like a car in real life.
This is one of Siemens’ main bets in its industry software vertical: in a demonstration of the Simcenter PhysicsAI system, the company showed how automotive component manufacturer Magna used the systems to improve the development time of its designs.
The parts manufacturer used the technology in crash tests between vehicles. With the use of AI, a collision design analysis that took 14 hours to perform now takes just 10 seconds. “This represents a 5,000-fold multiplier. In practical terms, 5,000 different project alternatives can be analyzed in the same time it would previously take to analyze one project”, says Mummaneni.
Continues after advertising
In a recent conversation with journalists in Brazil, the global vice president of startups and venture capital at AWS, Jason Bennett, stated that the technology giant is looking more closely at .
In addition to optimizing the execution of tests and modeling, Siemens’ perspective is that the use of physical AI will also generate a reduction in the cost linked to computing in engineering, called CAE. The company presented another case, this time from an undisclosed aerospace sector, in which the use of an artificial intelligence-based simulation predictor reduced CAE on a specific project by 60%.
The technology, called HEEDS, analyzes different possibilities for interacting objects with the environment, such as an aircraft in flight. “Engineers just define the objectives and HEEDS investigates the possibilities. This is where AI also becomes a multiplier of creativity in engineering”, points out Siemens’ senior vice president, Jean-Claude Ercolanelli.
*the journalist traveled at the invitation of Siemens
