Session: 03-09-01: Data-Driven Innovation in Smart Product Design and Manufacturing

Paper Number: 166271

Towards Trustworthy Digital Twins for Platform Technologies to Advance Biomanufacturing 

Digital Twin (DT) technology is transforming the biomanufacturing sector by enabling real time process monitoring, predictive insights, and enhanced operational efficiency. As the biomanufacturing industry shifts toward platform based manufacturing techniques, designed to enhance modularity, scalability, and adaptability, DTs are playing a crucial role in not only providing a virtual replica of physical manufacturing systems but also optimizing them through continuous, data driven improvements. However, ensuring the trustworthiness of DTs and their decisions in a platform based biomanufacturing setting presents significant challenges.

Firstly, the accuracy of DT models relies on the quality of input data and the precision of underlying algorithms. Uncertainties that arise from various sources, including incomplete process models, incorrect planning, human error, sensor inaccuracies, data variability, and algorithmic approximations can compromise the reliability of DT driven decisions, ultimately impacting manufacturing outcomes. Therefore, uncertainty quantification (UQ) becomes a crucial aspect of DT implementation, ensuring that decision making processes remain robust and dependable. To build confidence in the DT systems, robust verification and validation (V&V) frameworks are essential. Verification ensures that DT models are correctly implemented according to their specifications, while Validation confirms that they accurately represent real world processes. A well defined verification, validation, and uncertainty quantification (VVUQ) framework is therefore critical for managing risks associated with uncertainty and enhancing the trustworthiness of DTs in platform based biomanufacturing.

Secondly, with the increasing integration of Artificial Intelligence (AI) into DTs, new challenges have emerged regarding explainability, robustness, and regulatory acceptance of the AI models. AI driven DTs leverage advanced machine learning (ML) algorithms to analyze large datasets, identify patterns, and optimize biomanufacturing processes in real time. However, many AI models function as “black boxes”, making it difficult to interpret their decision making processes, which raises concerns about transparency and accountability. In response to these challenges, the U.S. Food and Drug Administration (FDA) has recently introduced a draft guidance on AI credibility assessment, offering guidelines to ensure that AI driven systems are trustworthy, fair, and compliant with regulatory standards.

Given the growing adoption of DTs in platform based biomanufacturing, this paper explores the key challenges and opportunities associated with their implementation in the platform based biomanufacturing domain. Specific focus is given on VVUQ requirements for DTs, DT and AI model trustworthiness, and a detailed exploration of the FDA’s recent draft guidance on AI credibility assessment. By addressing these topics, this work aims to provide a structured roadmap for developing trustworthy and reliable DT frameworks, supporting the continued advancement of platform technologies in biomanufacturing.

Presenting Author: Vishnu Kumar Morgan State University

Presenting Author Biography: Dr. Vishnu Kumar is an Assistant Professor in the Department of Industrial and Systems Engineering. Dr. Kumar brings extensive industry experience, having held various roles at Zimmer Biomet, CNH Industrial, OTIS Elevators, and Bharat Petroleum Corporation Limited, India. He is also Six Sigma Green Belt certified. Dr. Kumar is actively engaged in research, with publications in peer-reviewed journals and conference proceedings such as ASME JCISE and ASME IMECE. His work has earned him several accolades, including the Smith/Kern Graduate Research Exhibition Award and the Industry Xchange Best Poster Award from Penn State.

Authors:

Vishnu Kumar Morgan State University
Vijay Srinivasan National Institute of Standards and Technology