Geminus.AI announces the launch of its first product; a predictive intelligence platform that has the power to transform how companies design and operate industrial products and processes. The Geminus Platform fuses adaptive AI with physics using multi-fidelity modeling to enable creation of predictive models that uniquely combine high accuracy and speed, fast updating, and quantified uncertainty.
The product allows users to leverage key information on hand, such as low- and high-fidelity simulation, as well as real-world data, to efficiently create high performing predictive models that can be used for a variety of use cases including process design and control optimization.
"Lam's interest in Geminus is driven by the potential to employ their hybrid modeling capability to better predict how our equipment would behave in high volume manufacturing and optimize designs accordingly," said Faran Nouri, a director of Geminus and VP at Lam Research, an investor in Geminus.
The investments an organization makes in modeling technology can be staggering and the risks associated with generating expensive, ineffective models are high. Compared to conventional physics-based alternatives, Geminus improves ROI with faster, more effective models, reducing the risks of bad decision making. It also provides another step toward enabling digital twins, with all the advantages they can offer.
"We believe that AI in its current form will struggle to deliver ROI in complex systems that cannot tolerate insufficient accuracy and dynamically change over time. For example, our models can power digital twins that predict the behavior of highly complex processes, and enable increased productivity, at a level that does not yet exist today," said Greg Fallon, CEO of Geminus.ai.