AI-generated from publicly available materials.Researchers in South Korea are pioneering the development of a hybrid digital twin that integrates mechanistic modeling with artificial intelligence, aiming to create fully autonomous laboratories that require minimal human oversight.
Led by Dong-Yup Lee, a professor at Sungkyunkwan University, the team is advancing a digital twin that stands apart from traditional models by being continuously connected to a multi-sensory system that gathers real-time data from bioreactors. This innovative approach utilizes mathematical models of Chinese Hamster Ovary (CHO) cells to predict cellular behavior under various conditions, enhancing the accuracy of bioprocessing operations.
Lee highlights that while mechanistic models are essential, they often lack the adaptability needed for real-time applications. To address this, the researchers have integrated data-driven AI, which enhances the system's capacity for adaptive control. They also employ explainable AI (XAI) to clarify the factors influencing process outcomes, enabling better decision-making for bioprocess performance improvements.
Despite significant progress—approximately 80% completion of the hybrid model—the team faces challenges in effectively linking these models with control systems and future robotic applications. Looking ahead, Lee emphasizes that their research will focus on achieving greater autonomy in bioprocess operations, while also expressing interest in collaborations with industry partners to further advance this technology.