Title: Data sources for bioreactor digital twin development

Biography:

Bill Whitford is Strategic Solutions Leader, BioProcess, GE Healthcare in Logan, UT with over 20 years’ experience in biotechnology product and process development. He joined the company 16 years ago as a team leader in R&D developing products supporting biomass expansion, protein expression and virus secretion in mammalian and invertebrate cell lines. Products he has commercialized include defined and animal product-free hybridoma media, fed-batch supplements, and aqueous lipid dispersions. An invited lecturer at international conferences, Bill has published over 250 articles, book chapters and patents in several areas of bioproduction. He now enjoys such industry activities as serving on the editorial advisory board for BioProcess International.

Abstract

Benefits of incorporating digital biomanufacturing include increases in process understanding, monitoring and analytics; improvements in plant design and automation; better process design and flow; establishment of advanced embedded, distributed, or modular control units; and support of enterprise resource planning and manufacturing execution systems.  Key initiatives in the continued advancement of digital biomanufacturing in process development and control include de-siloing data, predictive simulations, model reference adaptive control, dynamic enterprise control algorithms and process automation.  The collection, organization and application of large amounts of data is a major element of this initiative.  Just one enabling technology here is termed “big data”. Generally, this refers to the use of very large amounts of structured and unstructured data from many sources in statistical and mathematical algorithms. In bioprocessing, sources of data include at-line or even on-line monitoring from both PD and manufacturing; early PD data from outside of the final design space; historic process data from validated production runs; process data from other sites running the same platform; or even data from disparate but related platforms.  Another important source is from such activities as published cellular platform metabolic pathways, flux and gene regulation-- along with general mammalian transcriptomic and proteomic understandings.  These data, coupled with well-designed algorithms provide both a quantitative understanding of cell physiology and advanced model-based control enabling advanced process development, operational efficiency, and business needs.  The ultimate goal is to generate individualized metabolic network and process parameter models to deeply represent specific host cell lines and bioprocesses.  Yet, even early Digital Twin designs are capable of creating predictive simulations that can improve process development, control and event prediction.