Abstract
The lifecycle of a medical mechatronic product, such as a CT scanner or MRI scanner, begins with the concept phase, which includes requirements engineering and model-based system development. This leads to detailed design in areas such as mechanics, electronics, embedded software, and operating system software. After integration, verification, and validation, the product moves into manufacturing, user utilization, and eventually refurbishment and recycling.
Throughout this lifecycle, extensive data in various formats are generated and stored in different data silos. For cross-domain data-driven decision-making that supports engineers across all areas, it is necessary to integrate this data into a cross-domain ontology. This ontology, as a "single point of truth," can be used for AI-supported assistance systems in various applications, such as embedded software code generation, PLC program creation, optimized production lines and layouts, and the definition of optimized manufacturing processes.
This presentation will discuss the complexity of the data, challenges of data integration, federated data management, and an ontology-based approach to data integration. Finally, examples will be shown on how these integrated data can be used to optimize LLM models (Large Language Models) using RAG (Retrieval-Augmented Generation) for various use cases.
Biographical Information
Dr.-Ing. Behrang Ashtari earned his master’s degree in electrical engineering from Leibniz University Hannover and completed his doctorate at the University of Stuttgart. His doctoral research, titled "Synchronization of Digital Twin with Real Production Systems," introduced the innovative Anchor-Point-Method—a method for detecting changes in real production systems and automatically updating corresponding Digital Twin models.
Since 2015, Ashtari has actively contributed to research and publications on Digital Twin technology and AI integration, particularly within the automotive industry, the German Aerospace Center (DLR) and Healthcare. In 2020, he joined Siemens Healthineers, where he serves as a Solution Architect in the Integrated Process and Tool Landscape (TE PLE) department. In 2024, he was recognized as a Senior Key Expert Architecture for Data Semantics at Siemens Healthineers AG.
Ashtari's work focuses on the seamless Data integration and Data Control across the entire product and production lifecycle—from requirements engineering to modernization and recycling. He utilizes an enterprise ontology model for federated data management across PLM, ERP, and MES systems to support Data-driven Decision Making.