To enable new applications on drone platforms, novel sensors coupled with with machine learning algorithms are essential. Such sensors are, for example, event-based cameras. They are fast (~kHz speed) and resilient to the environment (e.g. poor lighting conditions). Current cameras already utilize small FPGAs to convert the event stream to common interface standards, such as USB. However, for our drone system already incorporates a powerful FPGA for processing, these conversions waste energy and increase processing latency.
- Electrical and Electronic Engineering
- Drones (PBL), FPGA (PBL), Neuromorphic (PBL), PCB Design (PBL), Semester Project
Polymer networks are made by cross-linking polymer chains at their ends by means of a chemical reaction. While the properties of used reactions are usually very well characterized for small molecules, little is known about how the presence of a polymer chain and its length affect this reaction. In this project, we aim to study this, mostly experimentally, but also including a theoretical approach. We propose to start with boronic ester chemistry, which has been already characterized in literature and in our lab. the reactants will be functionalized on linear PEG chains. We plan on studying both the thermodynamic and kinetic parameters.
- Characterisation of Macromolecules, Physical Chemistry of Macromolecules, Thermodynamics and Statistical Physics
- Bachelor Thesis, Master Thesis, Semester Project
To enable new applications on drone platforms, novel sensors coupled with with machine learning algorithms are essential. Such sensors are, for example, event-based cameras. They are fast (~kHz speed) and resilient to the environment (e.g. poor lighting conditions). To utilize the full potential of these cameras, neuromorphic computing is needed to process their output. Our FPGA-based drone platform is capable of operating such an event-based camera, and allows for efficient hardware acceleration of the computationally expensive neuromorphic algorithms.
- Electrical and Electronic Engineering
- Drones (PBL), FPGA (PBL), Master Thesis, Neuromorphic (PBL), Software (PBL)
Bone exhibits a remarkable ability to adapt its microstructure in response to mechanical and metabolic demands. This process involves a dynamic balance between bone-forming osteoblasts and bone-resorbing osteoclasts, with osteocytes playing a crucial role in signaling micro-mechanical cues. Disruptions in these mechanisms, as seen in conditions like postmenopausal osteoporosis, lead to decreased bone density and increased fracture risk. Sclerostin is pivotal in determining bone formation or resorption in response to mechanical stimuli and is targeted by FDA-approved osteoporosis medications. However, the drug's mechanism and its interaction with mechanical loading remain unclear. This project aims to investigate bone's response to sclerostin antibody treatment and uncover the cellular mechanisms governing bone adaptation using single-cell mechanomics cluster analysis.
- Computational Biology and Bioinformatics, Engineering and Technology, Mathematical Sciences, Medical and Health Sciences, Physics
- Bachelor Thesis, Internship, Master Thesis, Semester Project
In continual learning, deep learning models incrementally learn more classes or tasks over time. Doing so, they should not forget previously learned knowledge. This is a hard and active research problem. Making it even harder, we want the models to also estimate correct uncertainty. E.g., they should be highly uncertain about a new object type, but not uncertain about an object that they just learned correctly.
 Parisi et al., Lifelong learning with neural networks http://dx.doi.org/10.1016/j.neunet.2019.01.012
 Gawlikowski et al., Uncertainty in Deep Neural Networks http://arxiv.org/abs/2107.03342
- Engineering and Technology, Information, Computing and Communication Sciences
- Master Thesis, Semester Project
Rotation optimization occurs in camera calibration, kinematics, animation, and attitude setting of spacecraft. These optimization problems are generically nonconvex and difficult to solve, but some problem variations offer exact solutions in the low-noise and outlier-free setting. This project will investigate Harmonic Hierarchies of polynomials as a mechanism to perform global optimization of outlier-robust rotation alignment tasks. Harmonic Hierarchies utilize the properties of the rotation space and binary hypercube to produce a sequence of linear programs in increasing size, yielding a sequence of convergent upper-bounds and lower-bounds to the true alignment error. Extensions to this project can include multi-camera alignment (group averaging) and problems in inverse kinematics.
This project will be performed in collaboration with Lucas Slot (ETH Zurich: D-MATH) and Mauricio Velasco (Universidad Catolica del Uruguay).
- Engineering and Technology, Mathematics
- ETH Zurich (ETHZ), Master Thesis, Semester Project
In this project, co-supervised by Swissgrid, we will look into the problem of designing incentives for the procurement of voltage support services in the grid. The design of an efficient mechanism requires the use of tools from control theory, game theory, and optimization.
- Electrical Engineering
- Collaboration, Energy (IfA), Master Thesis
For image and video and novel sensors, such as event-based vision processing, projects often make use open source libraries (e.g. OpenCV) for image processing on CPUs and GPUs, enabling fast prototyping and proof of concepts. However, drone platforms require algorithms to execute as efficient as possible. This allows less weight and energy consumption of the computing element, enabling longer flight duration or higher payload. FPGAs offer a promising approach, however, video processing algorithms are typically executed on hand-crafted hardware accelerators. To leverage the open source algorithms and libraries, efficient conversion from C and C++ to HDL accelerators is needed.
- Electrical and Electronic Engineering
- FPGA (PBL), Master Thesis
Organ perfusion is a method by which blood and other fluids are oxygenated and pumped through organs including livers, kidneys, lungs and hearts in order to provide the organ with oxygen and nutrients. Various organ perfusion technologies are already in clinical use to improve organ preservation or even treat organs prior to transplantation. Furthermore, ex-vivo perfusion offers the unique opportunity to study whole organs as an isolated system. As perfusion durations are extended up to two weeks, perfusion devices have to maintain good perfusate quality. In the human body, blood is filtered in kidneys which remove mostly water soluble waste products and excrete them in form of urine. This function can be simulated with a hemodialysis filter and dialysate for ex-vivo perfused organs. Here, a semi-permeable membrane separates blood from an aqueous solution called dialysate. Waste products and toxins then diffuse through the membrane into the dialysate which is pumped into a waste container. The rate at which dialysate is pumped into the dialysis filter determines how fast products are removed from blood. Unfortunately, hemodialysis does not only remove unwanted waste products but also medication such as anticoagulants and antibiotics. Therefore, we need a better model that allows for calculation of mass transfer between blood and dialysate.
- Biology, Clinical Engineering
- Bachelor Thesis, ETH Zurich (ETHZ), Internship, Semester Project
Whereas most of the locomotives owned by the Swiss Federal Railways (SBB) are operated fully electrically, and only new electric trains are being bought, diesel locomotives still exist and are actively used by partner companies. With the goal of a more sustainable operation and reducing emissions, these locomotives need replacement. However, fully replacing the diesel locomotives is economically inefficient, as most of them are still far from their end of life, and/or are operated in areas with no complete overhead line coverage. Thus, a locomotive being able to operate independently of overhead lines is desired, while being cost efficient. One solution to these expectations would be to retrofit existing diesel locomotives with battery-electric drivetrains.
- Electrical and Electronic Engineering, Interdisciplinary Engineering
- Bachelor Thesis, Energy Science & Power (PBL), Master Thesis, Semester Project