Conference Agenda

Engineering design activities range from classical process design to product design and manufacturing. This EURECHA-sponsored meeting focuses on discussion about this important feature of the chemical engineering curriculum and is divided into two parts: (a) a panel discussion in which invited panelists will present their approaches to the teaching of process design and associated courses; (b) a hands-on workshop in which participants working in teams will develop master plans for selected courses, with each team being mentored by one of the panel members. The first part of the workshop is open to all participants. The active part of the workshop will be open only to the active participants who register in advance (limited to 60 participants in total).
Course participants are invited to review a short video describing how the flipped class is used to teach process systems engineering (PSE) subjects (recommended before attending ESCAPE-35): click here to watch the video

The workshop introduces basic optimization and constraint-based methods in metabolic engineering and the development of large-scale kinetic models with a view to integrate strain and process design. The workshop extends with an introduction to digital twins for bioreactors and the demonstration of physics-informed neural networks in process development and control applications.

From Systems and Control to Process Systems Engineering and Beyond

This plenary presentation reviews academic career of Wolfgang Marquardt. He will give a glance of his early scientific contributions during his qualification phase as young junior researcher which set the initial condition for his later work. The focus of the presentation will be on his scientific activities at RWTH Aachen University. The talk will outline his long-term research strategy and introduce various thematic lines of his multi-faceted work in process systems engineering. He will sketch major ideas and results on selected topics, in particular, computer-aided support of mathematical modelling and engineering design processes, the exploitation of adaptivity in numerical methods for monitoring, estimation and control of large-scale process systems, optimal control methods and software and its applications, conceptual design and synthesis of hybrid separation processes, and integrated product, reaction pathway and process design with applications to tailor-made fuels from biomass. He will also sketch his roles and experience in science policy and research management in the last third of his career. The talk concludes with a few reflections on his learnings and a glimpse into the future of the field.

Annual meeting of the Eurecha members.

Process integration for industry decarbonization: Enabling a shared database of ex-ante models of industrial processes, decarbonization technologies and energy systems featuring sustainability metrics

The mass adoption of process simulators has accelerated data generation for the process synthesis and optimization, which is reflected in growing rates of scientific publications. However, many models are constantly redeveloped due to the limited scientific collaboration and confidentiality issues. Poorly documented databases also hinder the sharing of models between public-funded researchers, and research data is often static and difficult to reproduce, hampering comparability and reliability. To address these challenges, the creation of an open-source database for ex-ante models is crucial, facilitating knowledge transfer and promoting industrial symbiosis. This work aims to enable the procedures and infrastructure needed to build a shared database of ex-ante models of relevant industrial processes, decarbonization technologies and energy systems, with clear definitions of model attributes, adopted syntax, sharing protocols, reporting standards and visualization tools. The set of industrial and technological blueprints summarizes key material and energy flows without revealing confidential data, and being helpful for assisting in energy audits, identifying inefficiency sources and improving energy integration practices. The database of processes and energy technology models ranges from simple equation-oriented modeling approaches to complex sequential modular simulations with detailed internal flows, featuring a well-established documentation and versioning control strategy. By gathering information from different actors such as academics, researchers, industry experts, policymakers, opinion leaders, and stakeholders, the models database is continuously being built, validated and maintained, aiming to guide long-term decisions for the energy transition and demonstrate the role of Process Integration (PI) in industry decarbonization. These activities are developed within the Task XXIV of the Technology Collaboration Program of the International Energy Agency, whose objective is to share not only models but also experiences that inform decision-makers about the opportunities for industrial decarbonization using renewable electricity, CO2 capture and sequestration, and biomass conversion routes. The shared models also help highlighting synergies, such as utilizing waste heat and materials in other energy systems to reduce energy imports and minimize the environmental impact. The confidentiality calls for data anonymization and improved methods that employ minimal sensitive information. Finally, continuous education programs are integral to the dissemination strategy, serving as a conduit for training skilled engineers that contribute to the ongoing evolution of the Task. It creates a feedback loop that enlarges and refines the project’s database and models. This iterative process ensures that the project remains relevant and up-to-date, continuously adapting to challenges and opportunities in the industrial decarbonization landscape.

Higher education curricula are changing rapidly, shifting to more ML/AI in open-source and open-data environments often requiring coding, while sacrificing essential statistical thinking skills like understanding variation in data or conducting strategic experimentation. However, the ultimate questions are: Will the next generation of professionals in Computer-Aided Process Engineering (CAPE) and other fields be prepared for a data-driven future? Will the students really learn how to learn from data, e.g. to understand and optimize a process or to drive innovation and develop a new high-quality product?

Surprisingly, while students get access to more and more data and see more analytics in action in their daily life, the skill gaps reported by JMP customers in chemical, pharma or biotech industries are increasing. In this 1-hour workshop we will discuss these gaps and recommend some remedy, exploring best practices in teaching multivariate thinking and real-world problem solving. The JMP software used in this session is freely available to students, instructors and academic researchers at jmp.com/student, and all demo content will be shared with the participants.

Transforming Pharmaceutical Development: The Power of AI and Predictive Modeling

Presenters: Niels Vandervoort (Director digital transformation J&J) and Geert Craessaerts (Director Data&Process engineering)

In today's rapidly evolving pharmaceutical landscape, J&J is presented with exciting opportunities alongside exciting challenges. As our new molecules grow in complexity and the development cycles for new medicines are much shorter, it is clear that we must adapt our ways of developing new medicines to thrive in this dynamic environment. We will need a paradigm shift from iterative insights where we learn from doing to predictive insights in which we can anticipate product & process behavior at scale or in the patient. Therefore, digital transformation is not only a technological revolution; it is a business transformation in which we rethink how we develop our products and how digital can be an enabler and a catalyst.

By harnessing the power of new & advanced digital technologies like AI and by using the predictive power of advanced modelling; we will be able to leverage historical insights much better, and we can kick start our product & process design and start from a much better place of knowledge. Additionally, when we can generate the right insights faster through predictive models or by executing the right experiments through model-driven DOE and experimental iterations; we will be able to make better product & process design decisions much faster in the development lifecycle. Pharmaceutical companies typically make use of a wide range of predictive modelling technologies, including traditional methods based on mechanics, data-driven techniques, and hybrid models. We adopt our modelling technology based on what we know about the science, the physics and the availability of data. The presentation will highlight several examples from drug substance and drug product manufacturing, showing how these flexible modelling technologies help us with better and faster designing the processes. There is a great momentum for the broader usage of AI in pharmaceutical development given the support from authorities like FDA. However when entering the AI process modelling world, we should emphasize that transparency and explainability in AI systems are crucial for gaining trust from stakeholders and meeting regulatory requirements.

This year, ESCAPE35 will be co-hosting together with IChemE’s CAPE SIG a panel discussion under the theme “Career Options within the CAPE landscape”. The panellists will feature a diverse group of well-respected CAPE individuals from industry and academia, across different career stages. The panel discussion will focus on career possibilities within industry and academia for individuals within the CAPE community, and a networking event afterwards.

Our group has recently developed CODA, an AI-based platform to map whole diseased and healthy organs and organisms in 3D and at single-cell resolution. CODA solves the challenge of imaging large volume samples, while preserving high spatial resolution. Through integration with other multi-omic approaches – such as spatial transcriptomics and proteomics - CODA allows for unprecedented cellular and molecular profiling of tissues. I will discuss the new biological insights into tumor onset and progression gained from the use of CODA, including ovarian and pancreatic cancers, and associated biomedical implications for early detection of cancer. I will also introduce our new deep learning-based optical flow tool InterpolAI, developed to virtually restore damaged tissues and replace missing input images for enhanced 3D imaging, such as MRI, 3D histology, ssTEM, and light-sheet microscopy. Finally, I will describe our efforts to make CODA accessible to the wider community.

Process Systems Engineering through the lens of biopharma: niche challenges or research questions of the future?

Pharmaceutical product and process development rely primarily on time- and cost- intensive experimentation. In recent years, the adaptation and deployment of computer-modelling tools in this space have been gaining increasing interest as means to inform, accelerate, and optimise the industrial workflow. Despite, however, the modernization of the day-to-day workflows, the sector is still challenged by endogenous and exogenous uncertainty that often impair the ability to deliver enough therapeutics on time. This is further highlighted in the case of new modalities, where product and process performance are still unknown. At the same time, manufacturers are pressured to deliver therapies at accessible prices, while also considering the sector’s environmental footprint.

In this talk, we will discuss how Process Systems Engineering (PSE) innovation can enable quantification of uncertainty across the whole product lifecycle, from manufacturing to distribution. We will explore how PSE innovation can be adapted and transferred to cater to the needs of next-generation biopharma, in an effort to increase the sector’s preparedness, responsiveness, and efficiency. The inherent, bi-directional relationship between manufacturing and supply chains will be discussed with a view to understanding how supply chain optimisation can be revisited to satisfy the needs of an ever-changing medical landscape. Lastly, we will discuss how open challenges in today’s biopharmaceutical manufacturing and supply chain operation can motivate PSE research for the future.

Functionary

• MSc. Jan Luyts, Lead Functionary Project, essenscia, BASF
• MSc. Geert Boogaerts, essenscia, KU Leuven

Functional safety (FS) – an essential cornerstone in the process safety (PS) of chemical processes – is often approached as a purely risk-based challenge although this has several shortcomings:

• There is a lack of understanding of crucial safety parameters and how they affect one another. Furthermore, although having a major impact on the classification and performance of FS, risk assessments and quantitative failure rates are often subjectively determined through external stakeholders.
• The primary goal of a safety instrumented system (SIS) is to ensure safety, which often conflicts with cybersecurity standards and guidelines. It is essential to implement robust cybersecurity measures in the process industry to (cyber)secure the safety systems.
• A significant uncertainty on cost-determining factors throughout the plant’s safety lifecycle is often observed, resulting in the general trend of mismatching the goals and the means of FS.
• In process industry, safety departments have traditionally focused on random failures even though systematic failures are significantly more common, and even dominant in the total number of dangerous failures. Systematic failures, which are associated with human and organizational factors (HOFs), can lead to unacceptable scenarios that can no longer be mitigated. However, they are researched only to a limited extent to date.

This strategic project aims to tackle these shortcomings by developing an integral methodology to FS design based on an enriched performance- and technology-based approach that will shift the industry-wide paradigm. This new method will even further improve the safety of production units at a lower cost but without tolerating compromises on the overall PS. A more rational and efficient approach will be essential to safeguard the license to operate (LTO) and boost the competitiveness of the Flemish chemical industry.

This year, ESCAPE35 will be co-hosting together with IChemE’s CAPE SIG a panel discussion under the theme “Career Options within the CAPE landscape”. The panellists will feature a diverse group of well-respected CAPE individuals from industry and academia, across different career stages. The panel discussion will focus on career possibilities within industry and academia for individuals within the CAPE community, and a networking event afterwards. Confirmed panelists: Roberto Abbiati – Roche Seyed Soheil Mansouri – DTU (possibly) Lauren Lee - UCL