Ongoing digitalization in process engineering leads to completely new requirements for chemical and biochemical processes, but offers huge potential for more flexible and sustainable production as well. In a globalized market the change in raw materials and an increasing supply of renewable energies require processes which can react to changing conditions within a very short time. Personalized products and shorter cycle times of new variants require flexible processes for small quantities, which can be built up and stopped quickly and inexpensively. Strict requirements from legislators and regulatory authorities make a deep understanding of processes according to a "Quality by Design" indispensible.
These enormous demands on modern production processes require a radical rethinking in the design and operation of reactors for biotechnology, pharmacy and chemistry. In the future, flexibility of processes is more important than "economy of scale". Reactors of the future have to be "smart", i.e. quickly and flexibly adaptable to changing raw material qualities, energy sources and individual demands. In addition, "smart" reactors will be able to independently identify problems in the process (e.g. increasing pressure losses due to blockages, hot spots due to dead volumes, incomplete reactions due to short-circuit flows, poor heat transfer through scaling and fouling) and react autonomously, e.g. through changed throughputs, adapted heat supply or removal, as well as geometrical changes of internal structures.
However, in order to be able to intervene directly and specifically in case of a suboptimal process behavior, a deep understanding of the process is essential which can be acquired by using predictive models and simulations, and ideally leads to a simultaneously running, virtual process (a digital twin). Fast and non-invasive measurement techniques with high resolution, data-driven validated cross-scale models and simulation methods, intelligent control mechanisms, and new, intelligently structured reactors are required.
