Master Projects
Please submit your topic of choice by filling out the form below:
Stability assessment of lighting poles based on natural frequency
Supervisors: Xudong Jian (), Prof. Dr. Eleni Chatzi
Content
This project aims ot develop a tool to test the stability of street lamps and sports field masts in Switzerland by means of a natural frequency measurement and a design calculation. In previous bachelor theses, a Python-based calculation tool was developed to verify the structural safety of lighting poles according to SN EN-40. Until now, the tool has primarily covered metallic poles. The goal of this thesis is to extend the calculation framework to include concrete lighting poles and to expand its applicability to a wider range of mast types and linear attachments. Furthermore, the tool should be enhanced with a user-friendly interface and an automatic reporting function.
Tasks
- Implement the structural verification of concrete poles in the existing Python program.
- Develop an intuitive graphical user interface for the calculation tool.
- Extend the calculation to include linear attachments (e.g. climbing ladders).
- Create an automated verification document in Excel summarizing the results of the calculation.
- Expand the calculation to cover additional mast types (e.g. bracket poles, tapered poles, cylindrical-conical poles, double-bracket poles, double-arm poles).
Suggested Courses: Structural Identification and Health Monitoring
Possible to select as: obligatorishce Projektarbeit, praktische Projektarbeit, forschungsbezogene Projektarbeit
Limitation of offerings: 2 students (group work is possible)
Augmented Reality-assisted Finite Element Analysis
Supervisors: Kostas Vlachas (), Prof. Dr. Eleni Chatzi
Content
Augmented Reality (AR) offers immersive visualization and intuitive interaction with complex engineering systems. This project aims to couple AR technology with Finite Element Analysis (FEA) to enable interactive visualization and model calibration within an augmented environment.
The first goal is to build a visualization pipeline that renders FEA results—such as deformations, mode shapes, or stress fields—within an AR environment using Unity. The second goal is to establish a two-way interface between AR and the underlying modeling software, allowing users to adjust parameters, trigger analyses, and observe the updated model response in real time.
The project combines AR rendering, real-time data exchange (e.g., Python–Unity link), and user interaction tools to achieve a human-in-the-loop modeling workflow. The resulting framework will demonstrate how immersive environments can enhance understanding and calibration of engineering models.
Expected outcomes of project:
- Visualization pipeline linking FEA or physics-based simulation results to AR environments
- Development of Unity-based virtual/AR scenes for rendering dynamic responses and deformations
- Real-time communication interface between AR and computational modelling tools (e.g., Python/MATLAB/SOFiSTiK/ABAQUS)
- Human-in-the-loop model calibration based on AR-based user interaction
- Implementation of user interaction controls (e.g., model parameter sliders, simulation triggers) inside AR
- Demonstration of the end-to-end workflow on a benchmark FEA model
Suggested Courses: Structural Dynamics or Method of Finite Elements (or similar), Programming for Engineers (or any Scientific Programming course)
Suggested Competencies: Python/MATLAB, Strong background in coding
Possible to select as: obligatorishce Projektarbeit, praktische Projektarbeit, forschungsbezogene Projektarbeit
Limitation of offerings: 2 students (group work is possible)
Further open for: master students with background of computer science, electrical engineering, mechanical engineering, civil engineering, or related.
Experimental analysis of tilt estimation using acceleration measurements
Supervisors: Dr. Giulia Aguzzi (Kistler Instrumente AG), Prof. Dr. Eleni Chatzi ()
Motivation
Tilt estimation is a crucial aspect in structural health monitoring, where precise angular measurements are useful to assess the status of a structure. Traditional tilt sensors, such as inclinometers, can be costly or prone to drift over time. Acceleration-based tilt estimation offers a cost-effective alternative, leveraging widely available accelerometers to determine inclination. However, real-world factors such as noise, vibrations, and dynamic motion introduce challenges that require experimental validation. This project aims to explore and refine acceleration-based tilt estimation through a hands-on experiment, evaluating their accuracy, reliability, and practical applicability in dynamic environments.
Your Task
The objective of this project is to experimentally evaluate the accuracy of tilt estimation, derived from acceleration measurements, by comparing it to direct inclinometer readings. The study involves data collection from both accelerometers and inclinometers positioned in close proximity, followed by tilt computation using a provided estimation algorithm. If necessary, improvements to the algorithm will need to be explored to enhance accuracy and reliability.
Detailed task:
- Sensor Data Collection: Record acceleration and inclinometer measurements in the lab
- Tilt Derivation: Compute tilt angles using acceleration data based on a provided estimation algorithm
- Performance Comparison: Analyze discrepancies between the derived tilt and inclinometer readings
- Algorithm Enhancement: Identify potential improvements and refine the estimation algorithm if needed
- Report Findings: Summarize results, highlighting accuracy, limitations, and recommendations for further improvements
Possible to select as: obligatorishce Projektarbeit, praktische Projektarbeit, forschungsbezogene Projektarbeit
Limitation of offerings: 2 students (group work is possible)
Further open for: master students with background in computer science, electrical engineering, mechanical engineering, or related.
Optimal inspection and intervention planning for maximizing life and minimizing environmental impact of road infrastructure
Supervisors: Dr. Yunus Emre Harmanci (, Empa), Lukas Kramer (Empa), Prof. Dr. Eleni Chatzi
Content
Optimal inspection and maintenance planning of road infrastructure is a challenging process, since decision-making needs to be made on the basis of uncertain observations of various origin. These uncertainties are linked to some level of uncertainty relating to varying environmental and loading conditions, modelling errors, inefficiency of the measuring system and so on. Although this issue spans over several disciplines, including social sciences, it is also quite commonly met within an engineering framework. The goal of this project is to build an optimal inspection/maintenance framework for different bridge materials and typologies, maximizing their service life and minimizing environmental im-pact, and address the challenges of infrastructure management. Thus, the student will work on some of the following objectives depending on their interest:
- Review established and state-of-the-art non-destructive-testing and rehabilitation methods and derive indications on which methods work better for which different bridge typologies/materials
- Utilize existing finite element models (in SoFiSTiK) of various bridge types to simulate various short-and long-term effects.
- Explore decision processes frameworks for modeling decision making in situations with partly random or decision-dependent outcomes.
Suggested Courses: Depending on interest: Method of Finite Elements (or similar) and/or Non Destructive Evaluation & Rehabilitation of Existing Structures
Suggested Competencies: Basic knowledge on MATLAB or Python
Possible to select as: obligatorishce Projektarbeit, praktische Projektarbeit, forschungsbezogene Projektarbeit
Limitation of offerings: 2 students (group work is possible)
Autonomous Robotic Inspection for Bridge Vibration Monitoring
Supervisors: Dr.Xudomg Jian (), Prof. Dr. Eleni Chatzi
Motivation
Vibration-based bridge monitoring plays an important role in bridge structural condition assessment. Fixed sensors are mature but costly and not easily redeployed for this purpose. A mobile robot that can navigate, stop at designated points, acquire vibration data, and continue its route enables scalable inspection across many bridge sites. This project will deliver such autonomy using GNSS-RTK for primary navigation and infrared sensing plus computer vision for obstacle detection and safe operation.
Your Task
Build and validate an autonomous control pipeline on an existing intelligent wheeled robot. The robot will follow GNSS waypoints, perform stop and go measurement routines, avoid obstacles using infrared distance sensing and computer vision (optional), and log synchronized position and vibration data.
Detailed tasks:
Hardware and Software Integration: Integrate GNSS RTK, camera, infrared sensors, and vibration logger on Raspberry Pi 5, and implement synchronized data acquisition and configuration scripts.
- GNSS-Based Navigation and Stop-and-Go Logic: Develop waypoint-following control using GNSS coordinates and design stop-and-go behaviour with adjustable stop duration and robust recovery from navigation errors.
- Obstacle Detection and Safety Layer: Implement obstacle detection using infrared sensors and computer vision (optional) for proximity alerts, safe stops, and path adjustments, and log all safety events.
- Experimental Validation: Conduct supervised autonomous runs on a nearby bridge (e.g., Einsteinbrücke on Hönggerberg campus) or a mock outdoor setup, measuring navigation accuracy, stop precision, obstacle avoidance reliability, and vibration sensing quality.
- Data Processing and Reporting: Develop scripts for trajectory and vibration visualization, summarize measurement data with quality indicators, and compile a concise technical report with a demonstration.
Suggested Competencies: Comfort with Linux and embedded development on Raspberry Pi; Programming in Python or C++; Basic robotics or control coursework; Advantageous: basic computer vision experience
Possible to select as: obligatorishce Projektarbeit, praktische Projektarbeit, forschungsbezogene Projektarbeit
Limitation of offerings: 2 students (group work is possible)
Further open for: master students with background in computer science, electrical engineering, mechanical engineering, or related.
Bayesian Physics-Informed Neural Networks forgLinear Structural Dynamics
Supervisors: Dr. Rui Zhang (), Prof. Dr. Eleni Chatzi
Motivation
Physics-Informed Neural Networks (PINNs) are increasingly used to identify structural parameters from vibration data by embedding physical laws directly into network training. However, standard PINNs are fully deterministic and often overfit noisy measurements, leading to unreliable parameter estimates and limited robustness. Bayesian PINNs (B-PINNs) address these limitations by introducing probability distributions over the neural network or its outputs and applying Bayesian inference (BI) to obtain a posterior that is consistent with both data and governing equations. This allows the quantification of aleatoric (data-driven) uncertainty and epistemic (model/parameter) uncertainty.
This project investigates B-PINNs for linear structural dynamic systems, including single- and multiple-degree-of-freedom oscillators and a linear Euler–Bernoulli beam. The goal is to understand when Bayesian approaches improve robustness, how they behave under noisy measurements, and how uncertainty propagates into identified structural parameters.
Detailed tasks:
- Generate synthetic, noisy vibration data for simple linear systems (SDOF, MDOF, and a linear Euler–Bernoulli beam.
- Build baseline PINNs and BI for parameter identification and dynamic response prediction.
- Extend the PINNs to B-PINNs using: Hamiltonian Monte Carlo (HMC) or variational inference method.
- Quantify uncertainty in estimated parameters and response.
Compare B-PINNs against standard PINNs and BI methods in terms of accuracy and efficiency, noise robustness, and uncertainty quality. - Summarize findings with recommendations for when B-PINNs should be preferred over standard PINNs and BI in linear structural dynamics.
Suggested Courses: Structural Identification and Health Monitoring, Machine Learning, Programming for Engineers (or any Scientific Programming course)
Suggested Competencies: Python and MATLAB; strong coding skills;familiarity with Bayesian methods and PINNs is helpful
Possible to select as: obligatorishce Projektarbeit, praktische Projektarbeit, forschungsbezogene Projektarbeit
Limitation of offerings: 2 students (group work is possible)
Further open for: master students with background in computer science, electrical engineering, mechanical engineering, or related.