Imperial College London

Dr. Zoltán Kis

Faculty of EngineeringDepartment of Chemical Engineering

Honorary Lecturer



z.kis10 Website




C506Roderic Hill BuildingSouth Kensington Campus





We are performing computational modelling to support both the development and operation of vaccine production technologies. In the light of the current COVID-19/SARS‐CoV‐2 pandemic, we are using our models to aid the production of vaccines for addressing this global emergency. 

The new rapid-response, small scale, high-capacity vaccine production platform technologies, e.g. the self-amplifying RNA platform, are capable of producing vaccines against both existing and new pathogens (e.g. COVID-19/SARS‐CoV‐2), rapidly and at large volumes. To support their development we are carrying out: (1) techno-economic modelling to optimize process performance and reduce costs and (2) bioprocess modelling for quality by design (QbD) to ensure the high quality of the product during process development and subsequent manufacturing.

The combination of these exciting biological platforms, with advanced bioprocess engineering solutions and with computational models, can substantially accelerate clinical development, production process development and large-volume manufacturing. This facilitates pandemic-scale response, which requires the manufacturing and supply of over a billion vaccine doses within months from the start of a pandemic.

Techno-economic modelling of emerging vaccine production platform technologies

show research

To stop the spread and damage of existing/known and new/unknown pathogenic outbreaks, and to meet changing immunization needs in developing countries, large amounts of vaccines are required rapidly and at low costs. To address this challenge, at the Future Vaccine Manufacturing Hub, we are developing the following 4 new transformative vaccine production technologies: (a) self-amplifying RNA vaccines, (b) customizable virus-like particle vaccines, (c) customizable outer membrane vesicle vaccines and (d) humanized yeast-produced vaccines. These technologies can facilitate rapid-response vaccine production at high volumes. However, these are new technologies and it is not yet established how to best implement these for large-volume manufacturing.

Zoltán is supporting the development of the above four vaccine production platform technologies by modelling and simulating the production processes in order to identify the most technologically feasible and economically viable production options. In his process models, Zoltán considers continuous production, process intensification, lowering upfront capital investments by down-sizing facilities as well as innovative process designs. Production processes are optimized by de-bottlenecking and by performing sensitivity analysis to reduce costs and improve process performance. Our work guides the techno-economically feasible implementation of these emerging vaccine platform technologies.

Bioprocess modelling for Quality by Design

show research

We are developing and using a new Quality by Design (QbD) framework which integrates mechanistic or data-driven bioprocess modelling, product quality risk assessment, prior knowledge, with experimental and clinical data. This QbD framework will accelerate the development of new vaccine production technologies and will potentially speed up the regulatory approval process for new vaccines. Moreover, this will also assure that high-quality products are consistently manufactured, reducing the risk of production batch failures. This approach can in principle be applied to other biopharmaceuticals as well, such as monoclonal antibodies or cell and gene therapy products.

This QbD framework follows an iterative development cycle, it builds on prior knowledge and incorporates experimental lab data, pre-clinical and clinical data, and real-time production process analytics data, as these become available. The iterative development cycle includes the following steps: (1) identifying the patient needs (2) defining the Quality Target Product Profile (QTPP), (3) determining the critical quality attributes (CQAs) of the product and the CQA ranges using a risk assessment based approach, (4) determining the critical process parameters (CPPs) which affect the CQAs and determining the CPP ranges, (5) mathematically relating CPPs with CQAs, (6) based on the obtained model determining the design space and therein the normal operating range (NOR), (7) adapting the bioprocess model for advanced production process control, using model predictive control and real-time measurement data from the production process. This predictive model can forecast CQA values in the following time frame (e.g. next 10 minutes) and if CQA values are forecasted to violate the specified ranges, corrective measures, i.e. control actions, can be taken by the model, this way, preventing faults in product CQAs before these occur. Operating the process in the NOR offers the flexibility of modifying operating parameters in the GMP production process, instead of “freezing” the GMP process. This allows for production process optimization to overcome issues caused by the inherent biological heterogeneity.

This QbD framework which integrates computational modelling, risk assessment, prior knowledge, experimental and clinical data, can speed up the development and optimise the operation of vaccine production processes. This QbD framework follows an iterative development cycle to ensure continuous improvement through the product-process life cycle.


show research

Development and optimisation of technologies and production processes for the manufacturing of biotherapeutics (e.g. monoclonal antibodies, cell and gene therapy products).

Supply chain modelling and optimisation for biopharmaceuticals.

Assessment and modelling of industrial processes for improving technical feasibility, economic viability and environmental sustainability.

Research Student Supervision

Asma Zafar,, Development of a gene network for studying mechanosensing in endothelial cells. 2013-2014

Behmer,C, Bioprocess modelling for Quality by Design. 2019

Grace Freke,, Evaluating the activation of shear stress sensors by pulling with magnetic tweezers. 2013-2014

Karpouza,L, Techno-economic modelling of vaccine production processes in glycoengineered yeast. 2019

Oliver Fleck,, Evaluating shear stress sensors under various shear stress regimes. 2013-2014

Vanoutryve,J, Bioprocess simulation and techno-economic modelling of GMMA vaccines. 2019

Zhangxing Lai,, Design and construction of a flow chamber in which shear stress varies linearly along the axis of flow. 2012

van De Berg,D, Bioprocess modelling for Quality by Design. 2019