ERC Starting Grants for Four TU Delft researchers
The European Research Council has awarded four ERC Starting Grants to TU Delft researchers. The grants (1,5 million euros for a five year programme) are intended to support scientists who are in the early stages of their career and have already produced excellent supervised work.
The four ERC grants cover a wide variety of themes. Liedeweij Laan, for instance, will study how adaptive mutations improve fitness in yeast cells, in order to gain a better understanding of how organisms evolve. Wilson Smith plans to expand his research on electrochemistry as a means to transform CO2 and water to valuable chemicals and fuels, which may become essential in the near future due to the energy and environmental challenges humanity faces.
The project headed by Monique van der Veen is aimed at designing ferroelectrics by obtaining a fundamental understanding of how ferro- and piezoelectricity are related to a materials’ structure. Finally, Manuel Mazo Jr. intends to reduce the implementation and maintenance costs of so-called Cyber-Physical-System, which are digital systems that regulate and control many complex physical processes, such as chemical reactors or power networks.
Please read on for a more in-depth explanation of the four awarded ERC-proposals.
How do organisms evolve? How are proteins essential in one species lost in closely related species? Liedewij Laan will study how biochemical networks reorganize during evolution without compromising fitness. How organisms evolve is not only a fascinating fundamental question, but it also has important implications for human health: cancer and antibiotic resistance are poorly understood evolutionary processes. It is also a rather challenging problem: it is hard to know if a mutation has increased fitness, because this depends on the environment it arose in, which is typically unknown. Additionally, it is hard to find out how adaptive mutations improve fitness, because in cells all biochemical networks are connected.
Liedewij Laan will study how adaptive mutations improve fitness in yeast cells. She will use fluorescent live-cell microscopy in combination with physical modelling. Liedewij Laan will also create minimal evolvable in vitro networks consisting of emulsion droplets containing only the components that are essential for either fitness or evolvability, to find basic rules of network evolution.
More information about Liedewij Laan: http://laanlab.tudelft.nl/
The recycling of CO2 will play an important role in mitigating the energy and environmental problems that our future societies will no doubt face. Electrochemistry is a powerful technology that can make use of renewable electricity from solar and wind to power the sustainable transformation of CO2 and water to valuable chemicals and fuels.
The reduction of CO2, however, is not ready for large-scale deployment due to the poor activity and selectivity of catalysts that are currently used. New strategies are needed to improve our understanding of the complicated reaction mechanisms in order to gain better control of the electrocatalytic process. In this project, Wilson Smith will use a variety of operando characterization techniques and computational methods to gain fundamental insights into the reaction pathway of electrochemical CO2-reduction. He will receive an additional 500.000 euros to buy a specialised piece of equipment that can monitor the electrochemical process in situ.
More information about Wilson Smith: http://www.smithsolarlab.com/
Ferroelectrics can store and switch their polarity, and as such can be used as memories. They can also harvest mechanical vibrations via the piezoelectric effect. In this project, new ferroelectrics will be developed based on metal-organic frameworks, which may be used as physically flexible memories and mechanical energy harvesters for biocompatible sensors and implantable monitoring devices.
The materials most compatible with flexible substrates are soft matter materials. However, as energy harvesters, soft matter materials are hampered by low piezoelectric coefficients. The main objective of this research project is the rational design of ferroelectrics by obtaining a fundamental understanding of how ferro- and piezoelectricity are related to the materials’ structure, which can lead to materials with exceptional performance.
More information about Monique van der Veen: http://cheme.nl/ce/people/monique-van-der-veen/
As a result of the advances in electronic communication and computation, digital systems that regulate and control all sorts of physical processes, called Cyber-Physical-Systems (CPS’s), have become ubiquitous. Chemical reactors, water distribution and power networks are a few examples of areas in which CPS’s are used.
Systems such as this require the timely communication of sensor measurements and control actions in order to work as intended. Event-triggered control (ETC) techniques, which communicate only when needed, have attracted attention as a means to reduce communication traffic and save energy on so-called (wireless) networked control systems (NCS). However, the scheduling of the aperiodic and largely unpredictable traffic that ETC generates remains widely unaddressed – hindering its true potential for energy and bandwidth savings.
This research project aims to create models for ETC’s communication traffic and schedulers based on these models in order to increase the energy efficiency of wireless NCS by orders of magnitude. This will significantly reduce the implementation and maintenance costs of CPS’s.
More information about Manuel Mazo Jr.: http://www.mmazojr.net/Manuel_Mazo_Jr/Home.html
** Edit 15/3/2018 – Added following the arrival of Frans Oliehoek at TU Delft **
INFLUENCE – Frans Oliehoek (EWI)
Computers that can play Atari games or beat the world champion in the game of Go have sparked a renewed interest in the field of decision-theoretic sequential decision making (SDM). This field is concerned with endowing so-called intelligent agents, with the capability to choose actions that optimize task performance.
SDM techniques have the potential to revolutionize many aspects of society. Fundamental problems of scalability, however, prevent them from addressing certain complex problems. In order to overcome this barrier, Frans Oliehoek will develop a new class of influence-based SDM methods that overcome scalability issues for such problems by using novel ways of abstraction.
For instance, in the case of controlling traffic lights in an entire city, an intersection’s local problem is manageable, but the influence that the rest of the network exerts on it is complex. The key idea is that by using (deep) machine learning methods, sufficiently accurate representations of such influence can be made in order to facilitate near-optimal decisions. This project aims to develop novel decision making methods and demonstrate their scalability on two simulated challenges: traffic lights control in an entire city and robotic order picking in a large-scale autonomous warehouse.