The United States Patent and Trademark Office (USPTO) has issued patent US12585936B2 to Siemens Aktiengesellschaft for a novel method of executing machine learning models within industrial controllers. The patent describes a system where a controller's processor manages input values in a process image, and a co-processor applies a machine learning model to these values, writing the results back to the process image for use by application programs. This grant signifies a new intellectual property in the field of AI-driven industrial control systems.

While this is a patent grant and not a regulatory rule imposing direct compliance obligations, it represents a technological advancement that may influence future industry standards and product development. Companies operating in the industrial automation and AI sectors, particularly those developing or utilizing embedded machine learning in control systems, should be aware of this patented technology. No immediate compliance actions are required, but understanding the scope of this patent is crucial for R&D and product strategy to avoid potential infringement.