Theses
We offer theses and research internships in the areas of:
- Bayesian Optimization and Active Learning - used both for the creation of machine-learned (ML) Molecular-Dyamics-potentials for novel materials such as perovskites or Metal-Organic-Frameworks and for guiding synthesis in the labs of collaborators. (Milan Harth, Ioannis Kouroudis, Poonam)
- Graph Neural Networks (GNN) - creating and validating state-of-the art GNNs for usage as as surrogate model with data, produced with atomistic simulations at our chair: ranging from predicting catalysis-performance to solar-cell efficiency. (Poonam, Ioannis Kouroudis, Gohar Siddiqui)
- providing tools to our partners to solve typical data-wrangling tasks found in a modern, digital-data lab - including state-of-the-art ML tooling for image analysis (Ioannis Kouroudis, Milan Harth, Íñigo Iribarren Aguirre)
- advancing the performance of free-energy simulations incorporating large framework structures using CVs derived from ML (Gohar Siddiqui)
- extending our C++ based kinetic Monte Carlo for organic semiconductors (for example towards OLEDs) and electrochemical reactions (currently not active, maybe Gohar Siddiqui if you want to revive this in a Masters project)
- tooling for running complex simulation workflows: ranging from standardizing simulation protocols to providing tailored tools for 3D visualization (Íñigo Iribarren Aguirre, Gohar Siddiqui)
- if a economics/EDU-student reached here: improving our courses with automation ranging from thorough statistical evaluation/dashboards for student progress tracking to reviewing/prototyping options for plagiarism checks (Milan Harth)
In general all of these topics require a certain command of a classical UNIX-style shell environment and the willingness to learn how to setup complex computer experiments within this. This is something you are expected to pick up in bachelors thesis/internship or bring with you for a masters. In addition, at least basic command of a programming language is necessary if you want to work at your chair. For most ML-topics this is still Python (where we offer a course as module EI04024) which also is useful "glue" for any simulation/automation project and for getting the figures done in your thesis!
As you will be participating in our research effort, which is continously progressing, we don't offer more specific, fixed topics anymore. Instead we accept your application consisting of a CV, a transcript and a ranked list of 3 topics given above. We will then send anyone from TUM-EE a small take-home-test (about 1-2h of work) to judge your ability to work within the given problem areas. You have about a week to complete this assignment (soft deadline) and send us your work (code/diagrams/... even an outline of how you would tackle the problem). In a final interview you'll then explain your work.
Application: send your application to theses.sne@nano.cit.tum.de
Then you get your assignment(s) and will decide, whether we'll progress with an interview.
During the thesis work, we expect:
- participation in our weekly seminar. Either do a talk on the challenges you face with your problem or participate in the actual scientific discussion we do.
- if you are in Munich: seeing you in the student room at least 1 day/week, when your advisor is there.
P.S.: we have collaborators in physics, mech-engineering and chemistry who might co-supervise a thesis too. So if you're interested in our domain, don't hesitate to apply too, this needs coordination though, so no guarantees.
Current Open Positions
- MA_Machine_Learning_for_Tracking_Dynamical_Phenomena_on_Atomistic_Resolved_STM_Experiments.pdf
- MA_Machine_Learning_for_Boosting_Fast_STM_Images_Analysis.pdf
- MA_Creating_a_Database_on_Perovskite_Photovoltaic_Fabrication_by_Leveraging_Recent_Advances_in_Large_Language_Models.pdf
- Open_Doctoral_Candidat_at_TUM.pdf
Completed Theses
- Implementation of a Transfer-Matrix Method for Stratified Media in Modern C++
BA: dated 09/2022, Supervisor: Alessio Gagliardi, Manuel Gößwein - Artificial Intelligence for Image Recognition and Properties Extraction
BA: dated 09/2022, Supervisor: Alessio Gagliardi, Ioannis Kouroudis - Machine learning for the prediction of critical temperature of superconductors
MA: dated 08/2022, Supervisor: Alessio Gagliardi - Software-Entwicklung für Automobil-Test-Automatisierung und Tooling Anwendungen im Bereich Fahrer-Assistenzsyteme
IP: dated 05/2022, Supervisor: Alessio Gagliardi, Felix Mayr - Quantum Dot Property Prediction Using Bayesian Neural Networks
IP: dated 04/2022, Supervisor: Alessio Gagliardi, Ioannis Kouroudis - Machine Learning aided time series prediction for calcium concentration in cells
BA: dated 03/2022, Supervisor: Alessio Gagliardi, Ioannis Kouroudis - Feature dimensionality reduction of organic molecule representations for Machine Learning applications
BA: dated 03/2022, Supervisor: Alessio Gagliardi, Michael Rinderle - Prediction of Properties of Rubber Compounds in Tire Industry by Machine Learning Models
FP: dated 02/2022, Supervisor: Alessio Gagliardi, Ioannis Kouroudis - Erstellung und Entwicklung einer webbasierten Software zur Auswertung von UV/VIS-Messdaten von Mikroplatten-Readern
IP: dated 02/2022, Supervisor: Alessio Gagliardi, Felix Mayr - A differentiable framework for surrogate models of periodic systems in PyTorch
BA: dated 02/2022, Supervisor: Alessio Gagliardi, Ioannis Kouroudis - Computing Charge Transfer Integrals for Polymers - Using Fragment Orbital Density Functional Theory
MA: dated 11/2021, Supervisor: Alessio Gagliardi, Michael Rinderle - Testautomatisierung für Fahrer-Assistenzsysteme (Python Dev, API, GUI)
IP: dated 10/2021, Supervisor: Alessio Gagliardi, Felix Mayr - Creation and Benchmarking of a customized TorchANI model
Machine Learning Energy Potentials and Molecular Dynamics
BA: dated 09/2021, Supervisor: Alessio Gagliardi, Felix Mayr - Compute the Electronic Coupling of Organic Semiconductor Molecules using Constrained Density Functional Theory
BA: dated 06/2021, Supervisor: Alessio Gagliardi, Michael Rinderle - DFT-based computation of the transfer integral between organic semiconductor molecules
BA: dated 05/2021, Supervisor: Alessio Gagliardi, Felix Mayr - Calculation of Electrostatic Interactions in 2D Slab Geometries via MMM2D
BA: dated 03/2021, Supervisor: Alessio Gagliardi, Manuel Gößwein - Transition Metal Perovskites as Bifunctional ORR and OER Electrocatalysts Screened by DFT and Machine Learning
MA: dated 02/2021, Supervisor: Alessio Gagliardi, Felix Mayr - Machine Learning for the Investigation of 2D-perovskite Energy Band Gaps
MA: dated 01/2021, Supervisor: Alessio Gagliardi, Felix Mayr - Simulation of Electrostatic Interactions via Particle-Particle Particle-Mesh and Ewald Summation
BA: dated 11/2020, Supervisor: Alessio Gagliardi, Manuel Gößwein - Kinetic Monte Carlo Simulation of the Organic-Aqueous Electrolyte Interface
A first step towards elucidating charge transfer mechanisms
MA: dated 01/2020, Supervisor: Alessio Gagliardi, Waldemar Kaiser, Marlon Rück - Simulation of Core-Shell Nanowires for Thermoelectric Applications
MA: dated 01/2020, Supervisor: Alessio Gagliardi, Waldemar Kaiser - Comparison of Machine Learning Algorithms for IoT Applications on Low Power Microcontrollers
BA: dated 12/2019, Supervisor: Alessio Gagliardi, Michael Rinderle - Kinetic Monte Carlo Simulation of the Organic/Aqueous Electrolyte Interface: A first step towards elucidating charge transfer mechanisms
MA: dated 11/2019, Supervisor: Alessio Gagliardi, Marlon Rück - Kinetic Monte Carlo Simlation for Oxygen Reduction Reaction in Proton-Exchange Membrane Fuel Cells
MA: dated 10/2019, Supervisor: Alessio Gagliardi, Marlon Rück - Kinetic Monte Carlo Simulation of Filament Formation in Memristive Devices
MA: dated 08/2019, Supervisor: Alessio Gagliardi, Michael Rinderle - A Kinetic Monte Carlo Model for the Oxygen Reduction Reaction in Proton-Exchange Membrane Fuel Cells
FP: dated 04/2019, Supervisor: Alessio Gagliardi, Marlon Rück - Simulation of Core-Shell Nanowires
FP: dated 04/2019, Supervisor: Alessio Gagliardi, Waldemar Kaiser - Injection Models for Kinetic Monte Carlo modelling of Bulk-Heterojunction Organic Solar Cells
FP: dated 03/2019, Supervisor: Alessio Gagliardi, Waldemar Kaiser - Modelling of Recombination Losses in Organic Solar Cells
BA: dated 03/2019, Supervisor: Alessio Gagliardi, Waldemar Kaiser - Industrial placement at OPV applications, Merck
FP: dated 11/2018, Supervisor: Alessio Gagliardi - Semi-automatic solution for photoelectric-sensor testing-rail
IP: dated 10/2018, Supervisor: Alessio Gagliardi - Simulation of Core-Shell Nanowires
FP: dated 10/2018, Supervisor: Alessio Gagliardi, Waldemar Kaiser - Kinetic Monte Carlo Modeling of Exciton Dynamics in Organic Solar Cells
MA: dated 09/2018, Supervisor: Alessio Gagliardi - Machine Learning Bandgaps of Inorganic Lead-Free Mixed Halide Perovskites for Photovoltaic Applications
MA: dated 09/2018, Supervisor: Alessio Gagliardi, Aliaksandr Bandarenka - Multiscale Modelling of Charge Transport in Organic Semiconductors Utilizing the Concept of Machine Learning
MA: dated 06/2018, Supervisor: Alessio Gagliardi, Alexander Holleitner - Research Internship Report on Contact Models for Organic Solar Cells
FP: dated 03/2018, Supervisor: Alessio Gagliardi, Waldemar Kaiser - Polymer Chains in Kinetic Monte Carlo Simulation
BA: dated 11/2017, Supervisor: Alessio Gagliardi, Aliaksandr Bandarenka - MOVPE Growth and Characterization of GaInAsSb for the Integration in 4-Junction Solar Cells
MA: dated 11/2017, Supervisor: David Lackner EFTH, Alessio Gagliardi - Mapping of Trap Densities in Pentacene through Modelling and Simulation of 2D-Resistive Network
FP: dated 08/2017, Supervisor: Alessio Gagliardi, Mohammed Darwish - Charge Carrier Transport in Organic Semiconductor Devices: Establishing a Connection between kinetic Monte Carlo and Drift Diffusion Models
BA: dated 03/2017, Supervisor: Alessio Gagliardi, Tim Albes - Mapping of Trap Densities in Pentacene through Modelling and Simulation of 2D-Resistive Network
BA: dated 03/2017, Supervisor: Alessio Gagliardi, Mohammed Darwish - Modeling and Simulation of the Organophosphonate Molecules on Silicon/Aluminium Oxide Substrate by Use of the DFT
BA: dated 10/2016, Supervisor: Alessio Gagliardi, Mohammed Darwish - Visualization of Local Current Densities in Bulk-Heterojunction Organic Solar Cells
BA: dated 06/2016, Supervisor: Alessio Gagliardi, Tim Albes - Statistical Evaluation of Charge Carrier Dynamics in Organic Solar Cells
BA: dated 06/2016, Supervisor: Alessio Gagliardi, Tim Albes - Mesh Processing of Complex Morphologies in Solar Cells for Finite Element Simulation
MA: dated 11/2015, Supervisor: Alessio Gagliardi - Quantum Transport through monolayer of organophosphonate molecules on silicon/silicon oxide substrates
MA: dated 07/2015, Supervisor: Alessio Gagliardi - Validierungstests einer Motorsteuerung
IP: dated 02/2015, Supervisor: Alessio Gagliardi, Stephan Huber - Study on Perovskite Photodiodes
IP: dated 12/2014, Supervisor: Alessio Gagliardi, Aldo Di Carlo - Kinetic Monte Carlo Modelling of Bulk-Heterojunction Organic Solar Cells
DA: dated 05/2014, Supervisor: Alessio Gagliardi
MA: Master Thesis, BA: Bachelor Thesis, DA: Diploma Thesis,
FP: Forschungspraxis, IP: Ingenieurpraxis