WEBINAR (3-PART SERIES)

• End-to-end bioprocess digital twins [16 March]
• In medio stat virtus: Integration of mechanistic and data-driven (ANN) approaches in bioreactor hybrid modelling [1 April]
• Supporting journeys and enabling strategies for continuous downstream bioprocessing with science-based mechanistic models [27 April]

End-to-end bioprocess digital twins

Bioprocess digital twins are gaining traction in industry as part of a wider digital transformation. They are increasingly used for faster and more efficient process design and optimisation, risk assessment, scale up and tech transfer. However, the scope is often limited to a single unit operation considered in isolation from the rest of the process.

Whilst modelling a single unit operation at a time remains the most common application, gPROMS FormulatedProducts uniquely enables the creation of dynamic end-to-end bioprocess digital twins built from an extensive library of high-fidelity mechanistic models. Where mechanistic knowledge of the process is incomplete, hybrid approaches embedding data driven elements into mechanistic models can be used to fill the gaps.

In this webinar, we will provide an overview of the technical approach adopted at Siemens PSE for the creation and application of end-to-end bioprocess digital twins and discuss how they can provide value through a number of industrial case studies.

What this webinar covers

• An overview of industrial challenges and Siemens PSE’s vision on how end-to-end bioprocess models can support addressing them.
• Case studies showing where model-based solutions and the methodologies employed provide value to industry, including:
• Enabling R&D efficiency in whole bioprocess design and optimisation
• Assessing the impact of process disturbances and identifying failure modes to guide the process control strategy
• De-risking moving to continuous processes, technology transfer and scale-up of bioprocesses

Who should attend?

Those with an interest in bioprocesses from any industry, including:

• Non-modellers through to expert modellers in technical roles in R&D, engineering and operations.
• Decision-makers with an interest in learning about how bioprocess digital twins can bring value to their businesses across R&D, engineering, and manufacturing functions.
• Researchers and staff in academic organisations active in the bioprocessing field.

When

16 March 2022, 09:00 EDT/13:00 GMT

Duration

25 minutes plus Q&A

Presenter(s)

Dr Edward Close, Siemens Process Systems Engineering (SPSE)

Dr Edward Close (Edd) is the Practice Area Director for Bioprocessing based in the Formulated Products business unit at SPSE. He obtained his undergraduate degree in chemical engineering in 2009 and his EngD in biochemical engineering in 2013, both at University College London. His EngD was in collaboration with Pfizer and focused on developing mechanistic models of industrial purification processes to assist in process design and operation for the robust production of therapeutic proteins. Edd joined SPSE in 2013 and focuses on developing and applying SPSE’s solutions for upstream and downstream bioprocessing to deliver value to users across the Formulated Products industries.

In medio stat virtus: Integration of mechanistic and data-driven (ANN) approaches in bioreactor hybrid modelling

Bioreactor digital twins based on mechanistic models are increasingly getting attention in industry to help optimise critical process parameters while assuring process reproducibility, enabling process scale-up and technology transfer. The first principles approach enables extrapolation outside of the calibration space, facilitating transferability and scalability. Modelling upstream bioprocesses is challenging though, as the interaction between process conditions and living organisms is often poorly understood. As a result, it can be difficult to describe the entire design space with fully mechanistic models.

Data-driven models, such as artificial neural networks (ANNs), are an alternative modelling approach which can be highly predictive in the calibration space without requiring an understanding of the underlying mechanisms. However, with a purely data-driven approach, extrapolation beyond the calibration space is unreliable, the data requirements are high, and the uncritical execution of these techniques can easily breach physical principles and create minimal fundamental understanding. To address these challenges, a hybrid approach merging a data-driven element (e.g. ANN) within a mechanistic model is a promising solution to achieve the best of both worlds, increasing prediction capability whilst ensuring accordance with the physical laws and minimising data requirements.

In this webinar, we will discuss Siemens PSE’s vision for how hybrid bioreactor models can bring value addressing industrial challenges in upstream process development and operation. We will present a case study where the Mechanistic-ANN approach has been adopted to complement existing mechanistic knowledge.

What this webinar covers

• An overview of industrial challenges and SPSE’s vision of how hybrid model-based solutions can address them.
• A case study showing where hybrid model-based solutions for bioreactors and the methodologies employed provide value to the industry.

Who should attend?

Those with interest in bioreactor processes from any industry, including:

• Non-modellers through to expert modellers in technical roles in R&D, engineering and operations.
• Decision-makers interested in learning about how bioreactor digital twins can bring value to their businesses across R&D, engineering, and manufacturing functions.
• Researchers and staff in academic organisations active in the bioprocessing field.

When

1 April 2022, 09:00 EDT/ 14:00 BST

Duration

25 minutes plus Q&A

Presenter(s)

Dr Sarah Fadda, Siemens Process Systems Engineering (SPSE)

Dr Sarah Fadda is the Practice Director for Upstream Bioprocessing in the Formulated Products business unit at SPSE. She obtained her degree in chemical engineering in 2005 at the University of Cagliari (Italy) and her PhD in industrial chemistry and chemical engineering in 2009 at Milano Polytechnic (Italy). Before joining SPSE, Dr Fadda spent 8 years working as a research associate in academia. Her research focused mainly on modelling heterogeneity in various particulate systems, including biological applications, through population balances. Her last research project was at Imperial College London, where she gained experience in modelling biopharmaceuticals production systems. She joined PSE in 2017 and focuses on developing and applying SPSE’s mechanistic and hybrid solutions for upstream bioprocessing.

Supporting journeys and enabling strategies for continuous downstream bioprocessing with science-based mechanistic models

Separation using chromatography is key purification technique in the manufacture of many formulated products. Development of chromatography processes can be challenging as there is a demand to increase process productivity whilst maintaining product quality. Continuous chromatography processes have the potential to offer more productive processes as they allow for greater utilisation of the chromatography resins. However, development of optimal continuous chromatography processes is often challenging due to the need to conduct resource intensive chromatography experiments.

Science-based, data-calibrated digital twins derived from mechanistic models use process understanding combined with targeted experimentation to describe and predict the quantitative behaviour of processes for process design and optimisation, risk assessment, design space exploration, and scale up/tech transfer. The formulated products industries are increasingly using chromatography digital twins to address challenges described above and achieve these predictive goals, thus bringing value to their businesses across R&D, engineering, and manufacturing functions.

Examples of common challenges and questions we aim to address through using chromatography digital twins include:

• I am working on scale-up and optimisation. How can modelling speed up process development whilst providing a more fundamental understanding of the behaviour of a process?
• My organization is interested in continuous production to increase productivity – how can modelling help to switch from a batch to continuous chromatography processes?
• How can digital twins be deployed online to bring value in manufacturing?
• My organization has no/very little prior experience in mechanistic modelling. How can we harness the benefits of science-based digital twins without prior experience?

In this webinar, we will discuss Siemens PSE’s vision for how model-based solutions can bring value addressing industrial challenges in continuous chromatography process development and operation. We will then focus on case studies where model-based solutions have provided value to industry.

In these case studies, we will examine the methodologies used to deliver value, examples of quantification of this value, and the considerations made to select the model appropriate solution to solve the challenges presented.

What this webinar covers

• An overview of industrial challenges and PSE’s vision on how model-based solutions can support addressing them.
• Case studies showing where model-based solutions for chromatography and the methodologies employed provide value to industry, including:
• Enabling R&D efficiency in continuous chromatography process design and optimisation
• De-risking moving to continuous processes, technology transfer and scale-up of continuous chromatography processes
• PSE’s vision on how chromatography model-based solutions can bring value in manufacturing via the deployment of digital twin solutions.

Who should attend?

Those with an interest in chromatography processes from any industry, including:

• Non-modellers through to expert modellers in technical roles in R&D, engineering and operations.
• Decision-makers with an interest in learning about how chromatography digital twins can bring value to their businesses across R&D, engineering, and manufacturing functions.

When

27 April 2022, 09:00 EDT/14:00 BST

Duration

25 minutes plus Q&A

Presenter(s)

Dr Nehal Patel, Siemens Process Systems Engineering (SPSE)

Dr Nehal Patel is the Lead Developer for the chromatography libraries in the Formulated Products business unit at SPSE. He obtained his degree in chemical engineering in 2012 at Imperial College London and his EngD in biochemical engineering in 2018 at University College London. His EngD focussed on the use of mechanistic models and high-throughput experiments to aid process development. Before joining SPSE Nehal worked as a downstream development scientist for 3 years at GlaxoSmithKline where he helped develop purification processes for antibody-based therapeutics. Nehal joined SPSE in 2020 and focuses on developing and applying SPSE’s solutions for downstream bioprocessing.