Max Planck Institute for Dynamics of Complex Technical Systems

A new catalytic concept is laying the foundations for the methanation of carbon dioxide on an industrial scale

After two years of online only encounters, the Lindau Nobel Laureate Meeting 2022 took place onsite again

ContiVir, a Max Planck spin-off, presents a vaccine candidate against the Coronavirus, which can be produced in a quick, efficient and scalable process

Researchers make the proton pump of the respiratory chain work in an artificial polymer membrane

In order to immunize the world population both an effective vaccine and an efficient production process are necessary

In the event of an impending global flu pandemic, vaccine production could quickly reach its limits, as flu vaccines are still largely produced in embryonated chicken eggs. Udo Reichl, Director at the Max Planck Institute for Dynamics of Complex Technical Systems, and his colleagues have therefore been working on a fully automated method for production in cell cultures that could yield vaccines in large quantities in a crisis.

Normally, Peter Benner and his colleagues at the Max Planck Institute for Dynamics of Complex Technical Systems in Magdeburg work on complicated numerical methods to optimize the automatic control of technical systems and equipment. Recently, however, their research was applied to resolve a political conflict centering around drug cultivation, herbicide spraying and border violations in South America.

Wood waste and straw contain valuable substances for the chemical industry, and these substances are what chemists from the Max-Planck-Institut für Kohlenforschung in Mülheim an der Ruhr and the Max Planck Institute for Dynamics of Complex Technical Systems in Magdeburg want to get their hands on. The researchers are looking for ways to convert biomass into useful chemical compounds and use them as energy sources or raw materials.

Deubiquitinases are enzymes that regulate important signal transduction processes and protein degradation in the human organism. Mutations and dysregulation lead to an uncontrolled cell growth. Pathogens from bacteria and viruses are able to circumvent the human innate immune response system by mimicking the host’s intrinsic deubiquitinases.  Computer simulations are able to reveal mechanistic details of these processes and provide insight into the structural dynamics of activation and function of deubiquitinase enzymes in pathogens and human.   

We discuss the identification of nonlinear dynamical systems from data. Our approach is based on the symbiosis of operator inference and deep learning. Applications can be found, for example, in the design of digital twins in industry and technology.

A hybrid data-driven and mechanistic modeling approach is proposed for integrated material and process design. The method has been applied to a few example processes and substantial improvements on the process performance have been achieved.

In the age of electromobility, electrochemical energy converters such as fuel cells will play an increasingly important role in everyday life. On this point, diagnostic tools that can precisely determine the various fail states (flooding, drying out, catalyst degradation, poisoning, etc.) of these devices are becoming increasingly important. We report on a new experimental method for fuel cell diagnostics, based on frequency response analysis of concentration input and electrical output (current or cell potential), which can selectively distinguish between the different fail states.

Preparative chromatographic processes play an important role for the separation of complex mixtures. Recent research at the Max Planck Institute in Magdeburg aims at better understanding, systematic design as well as automatic control of these processes. Special focus in this contribution is on model based analysis of processes with implicit adsorption isotherms and a new self learning control concept for simulated moving bed processes.