EXPECTED RESULTS
Scientific and social impact
The objective of this scientific programme would be to build a model-theoretical framework to accelerate and improve decision making in the DBTL cycle.
The proposed research is rooted in advanced systems analytics and machine learning, knowledge and ontology engineering, mathematical optimization, multi-scale modelling, advanced process design, synthesis, and process
intensification. The research is ambitious introducing breakthroughs to connect underlying sciences (biology, metabolic engineering) of industrial biotechnology with lab experiments and process engineering.
Research capitalizes on rich, extensive, and diversified sets of data and knowledge available at the IBISBA infrastructure and platform.
IBISBA offers access to tools for in-silico computing experiments, extensive datasets from experimental labs and research pilots, facilities to validate innovations, as well as means to exploit and disseminate the research. The research will break with process development practices of the past leading to innovations that would reversecommunicate engineering questions back to science. One objective would be to upgrade computing and lab experiments into transferrable assets that could be shared across DBTL stages5. Challenges include the large volumes and the heterogeneity of data considering that they relate to diversified knowledge domains (substrates, enzymes, products, a variety of operating conditions etc.) rich in context. Such data are currently processed using simplistic, empirical, ad-hoc correlations that hold a limited predictive potential and are unable to yield phenomenological interpretations. DEBONAIR promises a coordinated application of machine learning technology with semantics organized using the TasCu2 (Fig. 4) ontology recently produced at IBISBA. A separate objective would be to connect computationally curated kinetics with kinetics from real-life experiments, essentially addressing a challenge to bridge methodological (knowledge) gaps between theoretical and practical methods. A final objective would be to develop embracing optimization models formulated over ensuing development stages (mostly between Design and Learn) so that to expand degrees of freedom in process engineering (letting the chemistry be a degree of freedom), develop adjoint maps, and reverse-engineer questions back to the science.
The research contributes to fundamental knowledge connecting the biorefinery concept with underpinning sciences, first principles and supporting its role as an integrator in bio-economy, industrial ecology and circular economy. The grant comfortably interlocks with my roles in ESFRI EU-IBISBA and ITN Marie Curie projects, adding value to all my research projects. In researching interactions with biological sciences, process engineering is poised to systematically exploit biological innovations bearing tremendous impact: enable the DBTL cycle to accelerate the development of biocatalysts, discoveries of generic paths to novel enzymes, debottlenecking engineering efficiencies, targeting substrates for products, opening new lines to explore highthroughput search across disciplinary borders in Industrial Biotechnology. Volumes of computational work is left unexploited due to existing gaps to translate underlying principles into practical recommendations. Advanced analytics and machine learning are generic methods to translate data across development stages and promise means to connect first principles with experimental work. Sharing standardized, clean, and consistent data across biorefineries is important. In connecting data to context and models, the research upgrades data to assets transferrable across experimental rigs. Research delivers new mathematical models to simultaneously optimize reaction pathways and process, as well as models to provide for the development of robust microbes resistant to space and time variations during the reaction. Above all, the developments are strongly connected to a most powerful consortium of practitioners and scientists able and keen to collaborate and contribute to the success of this effort.