A new system called FacProcessTwin has been developed to streamline the creation of process twins, which are real-time representations of production processes. This LLM-based system utilizes a large language model to generate process models from plant documentation and operator input, significantly reducing development time. In a case study with an Australian food manufacturer, FacProcessTwin achieved a 95.2% F1 score and built process twins in approximately one-sixth of the manual time, while also improving the accuracy of safety-critical bindings through human-in-the-loop governance. AI
IMPACT Facilitates faster and more accurate development of process twins, potentially improving efficiency in manufacturing.
RANK_REASON The cluster describes a research paper detailing a new system for process twin development. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →