Selected Fellows Call 3

High brilliance neutron source optimized for very high pulsed magnetic field experiments

Neutron scattering is one of the most powerful experimental tools for studying the structure and dynamics of materials on the nanometer scale. Among all, magnetic neutron scattering is a unique method which can be used to study magnetic structures of the materials which can lead to discovery of new kind of materials. However, there are two major limiting factors to magnetic neutron scattering: Neutron beams are expensive to produce, and therefore neutron data is statistically limited. Second, nowadays, high magnetic fields are achievable with pulsed magnets which produce a large heat load and therefore requires a long cool down period which means that such an experiment take a long time to perform.

In this project, we aim at improving the capabilities for neutron scattering in applied magnetic fields by synchronizing the neutron pulse with magnetic field pulses and design an optimized neutron source. We utilize a high energy proton beam and hit it to a heavy-element target and produce neutrons by spallation process. By optimizing the neutron source, i.e. the target, moderator, reflector assembly and transport the neutrons through the neutron guides, the goal is to achieve a higher brilliant neutron beam. The outcome would be radically new capabilities for neutron scattering under applied magnetic field.

Key words: Neutron Source, Spallatio, Pulsed high, High brilliance

Advanced peripheral neural interfaces

Peripheral nerve interfaces (PNI) have allowed remarkable achievements, from closed-loop control of prosthetic limbs to novel electroceutical therapies. However, the insertion step and the mechanical mismatch between the device and the nerve induce a Foreign Body Reaction (FBR) that hinders the chronic performance of the implant. Current peripheral nerve implants are fabricated using stiff silicon probes or flexible polymeric materials such as polyimide or Parylene, while the nerve is a soft yet tough, anisotropic, viscoelastic and non-linear tissue. I plan to fabricate a new category of implants to better match the mechanical properties of the host nerve. With the help of finite element modeling, I intend to identify the most suitable design to minimize mismatch. Moreover, I intend to fabricate devices that include soft materials such as hydrogels, since they mimic soft biological tissues and can be engineered to include antifouling properties that reduce the foreign body reaction. For each device I will fabricate, I will perform in-vitro and in-vivo experiments to assess the mechanical, electrical and biological performance of the design and material used. I plan to design implants for both stimulation and recording. Being part of the EPFLinnovators program, I will also carry out an internship in industry in the third year of my PhD. I plan to work on the integration of soft/flexible implants with signal conditioning and readout circuitry.

Key words: Neural interfaces, Soft bioelectronics, Peripheral nerves

Smart window systems for maximizing natural ventilation potential in European apartment buildings

Because of high building envelope airtightness and the heatwaves caused by climate change, the newly constructed and renovated European buildings are often facing indoor air quality (IAQ) and overheating problems, respectively. Previous studies demonstrated that adequately designed and applied Natural ventilation (NV) can solve the above-mentioned problems with a very low or no energy. Nevertheless, as the NV is based on natural forces and occupant behaviour, its use may not always be optimized. With the development of the Internet-of-Things and window automation technology, there are the tools to optimally exploit NV’s full potential, which was not available to date. The purpose of this thesis is to examine the maximum potential of NV in the European apartments and to enhance its performance concerning energy use and indoor environmental quality (IEQ) with the use of smart automated windows. This study deploys a combination of statistical analysis on ambient air pollution and climatic data, and building energy simulations (BES) in parallel with field experiments to examine with what degree of success the intelligent NV control systems can be implemented in apartment buildings and if they can eliminate the overheating risk and improve the IEQ with minimum cost and energy. The expected outcomes of this study are (i) the estimation of NV potential for the European apartment buildings and (ii) the development of an optimized, automated window operation for this building type.

Key words: Natural ventilation, Energy savings, Ventilative cooling, Window automation

Advanced ManUfacturing Solutions for active building Elements (AMUSE)

The increasing energy consumption and the necessity to reduce contaminant emissions has pushed governments in Europe to introduce initiatives such as Nearly Zero Energy Buildings (NZEB), which forecast an enormous adoption of renewables, especially Solar Photovoltaics (PV), into the built-environment. Besides the political thrust, one of the main drivers of PV integration into buildings is the development of more aesthetically pleasant products. Bringing PV into the built-environment leads to more complex manufacturing processes compared to standard methods due to high product customization (e.g. shapes and sizes). Therefore, manufacturers must rely on expensive and inefficient production processes to manufacture Building Integrated Photovoltaic (BIPV) elements. The masking of metallic ribbons, for instance, may increase the visual appeal of BIPV products. How to address this in manufacturing in an efficient way is however not straightforward and requires solving multiple problems. The goal of this thesis is to design a flexible and versatile ribbon coating equipment (RCE) as a proof of concept that could be prospectively integrated in a BIPV manufacturing line, aimed at masking the metallic ribbons of a solar module with ink. The reliability of the newly developed BIPV components will be assessed and understood. The introduction of the equipment into a manufacturing line of BIPV products will be analyzed in collaboration with the industrial partner of the project.

Key words: Building Integrated Photovoltaics, automation, coating, aesthetics

Conditions of habitability in a Contemporary Company Town. Rethinking the industrial cities and their productive habitats

Company towns appeared worldwide during the early industrial age to enhance the living conditions of production sites and their surroundings, maintaining the social control of employees through corporate welfare policy built around the concepts of efficiency, paternalism, and hierarchy. Nowadays, most of the company towns resumed their activities, but some of still exist. Focusing on the city of Dalmine in Northern Italy, founded in 1907 by a German company and still active in the steel pipe sector and known as Tenaris, this study aims to (1) understand how the relationship between man, labor and territory has changed in a company town context; (2) investigate how living conditions in a company town have changed over time; (3) examine if some patterns of the heritage of the company town persist until today and may influence its actual living conditions. By means of extensive fieldwork inside the industry supported by Fondazione Dalmine, the collection of cartographical, technical, and iconographical business archives, and qualitative methods with the local community, workers and authorities, the study applies a cross-cutting microhistorical and urban analysis approach to identify the remains of the modern industrial city and propose to recognize past and actual habitability. The study contributes to the research of industrial areas facing urban, technological, and environmental transitions and intents to conceive new approaches, policies and practices in areas deeply marked by the historical influence of the industry.

Key words: Company town, habitability, productive habitat 

Delivery of cancer neoantigen vaccine with a carrier-free nanogel for personalized immunotherapy

Two patients diagnosed with lung cancer, one successfully survived for a very long time after treatment. However, another patient who received the same treatment may suffered from multiple cancer recurrences. To improve the therapeutic outcome of immunotherapy, we start to develop personalized vaccine based on neoantigen. By using DNA sequencing to capture the specific features of individuals’ tumor, then interpreted these features into a short chain of amino acid, the peptide. This special peptide bearing individual tumor information can be used to fight against cancer.
We are developing a safe and effective peptide vaccine formulation, called the PEV vaccine. In a conventional vaccine formulation, the main functional components are the antigen and adjuvant, which can protect us from the diseases. In our work, we consider these components in vaccine formulation as the passengers. And our PEV can serve as the vehicle to protect these passengers from damage and efficiently transport them to the destination. With the help of our PEV, the working efficiency of antigen and adjuvant in vaccine formulation can be significantly improved, more effectively protecting us infection and diseases.

Key words: Personalized vaccine, immunotherapy, neoantigen

Targeting biophysical cues of tumors for enhanced immunotherapy

Cancer incidence and mortality are rapidly growing worldwide, and cancer is expected to be the leading cause of death and the most significant barrier to increased life expectancy in the 21st century. Today, a majority of patients are treated with conventional measures such as chemotherapy and radiation therapy, with limited effectiveness and severe side effects. The emergence of immunotherapies, a treatment that harnesses the patient’s own immune system to combat cancer, in the past decade has provided a dramatic shift in the fight against cancer. In particular, chimeric antigen receptor T cells (CAR-T) therapies have shown tremendous success in the treatment of blood cancers, such as leukemia and lymphoma. Unfortunately, to date T cell-based strategies have not yielded significant results in the treatment of solid tumors – one of the most daunting challenges in oncology. The tumor microenvironment (TME) in solid tumors presents several key challenges to T cells, including a strong immunosuppressive environment and the presence of multiple obstacles to tumor infiltration. To date, cancer immunotherapies have overwhelmingly focused on some of the biochemical or molecular hallmarks of cancer, with limited success in solid tumors. However, tumors exhibit a complex microenvironment with distinguishing biophysical and mechanical characteristics. Among others, tumor tissues are typically significantly stiffer than healthy tissues. Interestingly, tumor stiffness persists over long periods of time, displays higher homogeneity compared to protein and genetic biomarkers, and is quite prominent and universal across solid tumors. Such increased tissue stiffness represents a unique physical biomarker of cancer and has already been employed for the diagnosis of breast and prostate cancers in the clinic. On the other hand, immune cells are known to sense, respond, and adapt to the mechanical characteristics of their environment. We aim at leveraging the mechanosensing abilities of immune cells to target specific biophysical cues of solid tumors. This approach may improve specificity of immunotherapy while reducing off-target effects.

Key words: Immunoengineering, Cancer therapy, Mechanobiology

Towards assisting strategies for lower limb exoskeletons applied to stroke rehabilitation

Lower-limb exoskeletons (LLEs) hold a lot of promise in helping people overcome mobility impairments, as devices for continuous usage and rehabilitation. Control of exoskeletons is crucial to their functionality, and important challenges in the development of these devices are related to control. Effective interaction with the user is a serious challenge in control of LLEs for assistance. Taking interpersonal differences into account is important for effective interaction in general, and particularly in rehabilitation applications in which the optimal assistance needs to match each patient’s gait pattern, impairment and recovery level. However, most of the control methods for LLEs rely on generic trajectories or torque profiles. Even though strategies exist for adjusting the assistance to each individual, this is rarely at the core. Moreover, many of the potentially patient-tailored methods have not been properly tested for effectiveness. Another often-neglected aspect is assistance in transitory movements (e.g. sit/stand transitions) and alternative locomotion modes (e.g. stair ascent/descent). The focal point of this project will be control of LLEs for personalized assistance, with stroke patients as the target population. Particular challenges associated with stroke patients include their need for asymmetric assistance and not hindering their intact motor abilities. In addition to walking, assistance of sit/stand transitions and stair climbing will also be addressed.

Key words: Lower-limb exoskeletons, Stroke gait rehabilitation, Assistive control

Future Changes in Dynamics of Extreme Precipitation and Dry Periods in Swiss Alps and Consequences for Hydropower Production Optimization

As the frequency of extreme events in alpine regions is expected to increase due to climate change, understanding the spatio-temporal changes in the occurrences of these events is important for the security of energy derived from renewable resources, which are key to combating global warming. The main objective of this PhD work is to attempt to quantify the expectation of the frequency of extreme precipitation events and dry periods in the Swiss Alps, which would pose a risk to the water supply in reservoirs used for hydropower production. Traditional and novel statistical models for extreme value analysis (EVA) are used to provide an initial estimate of the return periods for high inflow events for an alpine reservoir in Ticino as a first case study. Given the extensive engineering in the Swiss Alps, it is often the case that reservoirs which may not be hydrologically connected naturally, supply water to the same turbine downstream. Therefore, a model is needed to translate precipitation events, and in particular their space-time dynamics, to runoff extremes for inflow (or lack thereof). In total, six hydropower systems with a combined sum of 11 basins/reservoirs will be analysed in this way. In addition to furthering understanding of extremes in the Alps, the results of this work will be tested as an input into existing optimisation schemes for hydropower production and trading.

Key words: Extreme events, extreme value analysis, spatial statistics, hydropower

Porous media microstructural design for the performance improvement of proton exchange membrane fuel cells

This thesis is focused on the understanding, modelling and experimental validation of the potential effects of ordered anisotropic microstructure porous media in the gas flow channel of PEMFC. This study reviews the components of protons exchange membrane fuel cells (PEMFCs), the fundamental working principle, and the hurdles toward commercialization. The main obstacles toward the mass production of PEMFCs are their cost, durability, and cold start abilities. , which are always interwined. Because thermal and water managements are crucial, efforts are concentrated on the development of numerical models and modern imaging techniques to predict the system behavior and to optimize the system design and operation. The governing equations to model the relevant physics in all the components of the system are also provided. Novel flow field designs for the bipolar plates (BP), or developing new microstructures for the gas diffusion layer (GDL) or microporous layer (MPL) are also linked to the topic of current thesis due to their physical coupling ot the gas flow channel. The relevant length and time scales varies over 10-50 nm orders of magnitude. PEMFC comprises components the functionality of which is provided by their 3-dimensional nature (for the transport of electron, ions, gas and liquid phase reactant, etc). Therefore, advanced imaging methods are required to understand how their microstructure must be designed to achieve target effective properties. Herein, FIB-SEM is adequate for the imaging of 1815 × 1400 × 1500 nm^3 and elemental mapping using Energy Dispersive X-Ray (EDX) analysis will be performed to determine the presence of different materials and components. The knowledge gained at the micro- nanoscale will allow to optimize the design, operation, and microstructural properties using artificial neural networks to develop a model based on the obtained numerical and experimental data.

Key words: Thermal/water management, Proton Exchange Membrane Fuel Cell (PEMFC), Porous media

Sense Dynamics

In this project, we envisage to develop a navigation system to ensure localization of a drone during GPS outage (or in general Global Navigation Satellite System – GNSS) by utilizing a vehicle dynamic model (VDM) and distributed sensor network. GNSS outage could be circumstantial (in mountains, forest etc.) or intentional (spoofing, jamming etc.). This sensors network shall quantify wind vector and its effects on the flying drone (which includes aerodynamic forces and moments) and pass on this information to the VDM based estimator. In order to achieve the aforementioned objectives, a noise characterization of several in-house and prospective air-velocity sensors shall be carried out using a Generalized Method of Wavelet Moments (GMWM). Secondly, it is envisaged to integrate a Pitot tube to VDM architecture to find out the best that can be achieved with an existing state of art sensor. Subsequently, the same exercise shall be repeated for a distributed sensor network to estimate the wind, its effects and the corresponding uncertainties. Having this information is believed to establish new state of art results for Beyond Visual Line of Sight (BVLOS) drone navigation.

Key words: Navigation, GNSS (GPS) outage, Vehicle Dynamic Model, Distributed sensor network

Development of a synthetic eye to accelerate the discovery of improved drugs for the treatment of retinoblastoma

Retinoblastoma is a type of eye cancer that affects children and may result in irreversible loss of vision or even enucleation of the eye. While current treatment strategies are highly effective and have been optimized to save both the eye and the vision, the side effects of the treatments remain problematic, and treatments that are more selective would be beneficial.

To improve treatment selectivity and decrease adverse effects, heat-activated drugs in combination with hyperthermia were proposed as a better option. At the moment, combination of chemotherapy and hyperthermia is standard-of-care treatment modality. However, the drugs used in the treatment were not designed for this particular application, so, presumably, better drugs could be discovered.

Finding potential drug candidates and treatment options through the use of the in vitro model based on human cancer organoids would contribute to the problem solution. It could accurately predict drug efficacy, would be a logical step between the cell culture and animal studies in the drug evaluation timeline and can be considered as a synthetic eye model.

Key words: Retinoblastoma, in vitro three-dimensional model, combination therapy

Methods for the sustainable classification of reusable building components in demolition buildings and their incorporation into the construction material cycle

The construction industry is one of the major sources of total greenhouse gas emissions that lead to climate change and environmental issues. It generated enormous quantities of waste every year through conversions and dismantling, amounting to 16.8 million tonnes throughout Switzerland in 2018, with an expected annual increase. Beyond that, the international construction sector is responsible for a significant part of raw materials’ consumption.

Such issues render the building sector increasingly dependent on resource-efficient reuse of building products. Every building is a reservoir of so-called “grey energy”, which could be used sustainably by being reused in new projects. Due to advances in the circular economy, recycling of building materials is constantly evolving, whereas reuse of building components is still challenging, as such elements are not available in large-scale retail sale and barely implemented into the built environment.

The doctoral thesis intends to verify the key obstacles for incorporation of reusable building components into the construction material cycle, point out methodical assessment and develop suitable instruments to significantly improve reuse of construction components. The research aims include the definition of key stakeholders and selection of building components (depending on the period of use, dismantling capacity and time value), according to their categorisation (material accessibility, type, quantity, size and mass) and options for further reuse.

Key words: Reusable building components, construction elements, reuse, dismantling

Building Climate Readiness

Increasingly frequent and severe weather events have highlighted the continued threat of climate change to the built environment as a secure asset class. As a result, there is a clear need for real estate investors to properly assess their vulnerability to climate risks to make the informed financial decisions necessary to protect their asset’s value.

Advances in remote sensing, image processing, and forecasting have the potential to improve the accuracy of automated valuation models. By incorporating geospatial techniques and partnering with Wüest Partner, a real estate consultancy, this research project aims to assess whether, and if to what extent, climate risks are already captured in current asset valuation. The motivation of this project is to develop a framework in which physical and transitional risks, as well as forward-looking property-level vulnerability or “climate readiness”, can be systematically learned and evaluated.

Key words: Real estate valuation, mass appraisal, asset price, climate risk

Microfluidic platform for high-throughput phenotyping and chemical compound testing using the nematode Caenorhabditis elegans

I aim to study the C. elegans nematode by means of microfluidic systems. Indeed, I spy on worms while they are cultured inside my microfluidic chips, by tracking them one by one in the long term. We have chosen this particular type of creatures as it has unique features, which make it a favorable target for the life sciences research. It is a soil worm with an approximate size of 1 mm in length. Its lifecycle is only in order of few days, which gives it a great potential for the life-long observation, especially in comparison to other animals commonly exploited in life sciences, such as mice or rats. In addition, it is relatively costless, and there is no ethical constrain. I develop an automated set-up to confine the worms inside the microfluidic chambers, feed them via an automated liquid delivery system, perform large-scale microscopy and acquire meaningful data. By establishing such a platform, I would be able to monitor the worm development in response to a wide range of parameters such as the feeding rate and quality, temperature, toxins and drugs (called phenotyping). C. elegans shares more than 50% of its genes with human, turning it a perfect choice for drug discovery, toxicity assay and gene therapy. Its small size and short lifetime allows high-content and high-throughput phenotyping, that is always demanded by biologists. My goal is to develop a reliable and precise platform to perform the above-mentioned tasks and finally gain new understandings.

Key words: Microfluidics, C. elegans, Microsystems

Occupant-centric sensing and control strategy for optimal air quality, comfort in office building

As people spend most of their time indoors, securing a good indoor environmental quality has become important in building sectors. Heating, Ventilation and Air-Conditioning (HVAC) system which is the biggest energy consumer in a building is strongly associated with occupants’ health, productivity, and comfort. There have been investigations to advance HVAC control strategy while introducing smart and efficient sensing and monitoring strategies. However, there has been limited investigation on ‘what is optimal sensing method under various environmental/contextual variations in the space?’. Specifically, a systematic guideline of what, how, and where to measure when it comes to monitoring indoor environments is still not defined. This study aims to derive advanced optimal/proxy sensing strategies/guidelines while considering human exposure to indoor air pollutants (CO2, PM, VOCs) and investigating spatial and temporal IAQ variations in various office contexts. This study will expand our understanding of not only optimal sensing technique itself but also its feasible and cost-effective implementation under a real dynamic context in office. Moreover, this study will contribute to improving the indoor environmental quality of building with advanced sensing strategy and HVAC control, which will ultimately improve occupants’ health and comfort.

Key words: Indoor air quality, Contextual sensing, Occupant-centric HVAC

Rational Design of Novel Multivalent Antivirals

Viral attachment inhibitors have been studied for decades in the effort to develop effective antivirals. As a class of newly discovered viral attachment inhibitors that bind viral receptors irreversibly and cause viral disruption, virucidal molecules are considered promising candidates for antiviral drug development. In SuNMIL, a new class of virucidal molecules have recently been developed, which share a common structural scaffold that features a rigid core attached with multiple viral attachment ligands through long flexible linkers. It has been suggested that the length, flexibility and hydrophobicity of the linkers are critical to the virucidal activity. However, the structure-activity relationship of virucidal molecules has never been systematically studied, mainly because the lack of model molecules able to address the outstanding questions. In this study, we will build a structurally defined virucidal compound library based on small-molecule cores and investigate the effect of structural parameters like linker length, number, hydrophilicity and rigidity, in order to better understand the interaction between viral receptors and virucidal compounds. Meanwhile, clinically important properties, such as solubility and toxicity, will also be systematically studied and optimized. We expect the final outcome of this project to be the development of an ideal broad-spectrum antiviral compound.

Key words: Broad-spectrum antivirals, multivalence, virucidal


Funded by
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 754354.