SA/BA/MA projects
Have you heard of industry 4.0, smart grids, smart buildings, smart cities, intelligent traffic systems and intelligent self-driving cars? Did you ever wonder what makes them smart and intelligent?
The answer is control and automation and that’s what we do at the Automatic Control Laboratory (IfA). We use control theory, optimization, machine learning, and game theory to develop controllers and algorithms that are the backbone of nearly all modern technology. At our lab we span the whole area from pure theory to real-world applications and we are looking for you to help us push forward the state of the art. If you want to learn techniques and gain knowledge that enable you to work in any field from medical applications to spacecraft and from electrical grids to finance then the Automatic Control Laboratory is the place for you!
Lists of currently running and recently completed projects can be found in the sub-navigation menu above. Reports for several projects are available through the ETHZ Research Collection:
Open projects
Optimal fertilizer application under imperfect ensemble weather forecasts using RL
This project combines algorithmic challenges in the design of machine learning based control algorithms with recent advances in precision agriculture. The student will implement, analyse and improve novel reinforcement learning approaches under partial state information. The algorithm will then be applied to achieve an optimal irrigation and fertilization scheme in precision agriculture under uncertain weather predictions. A simulator of the soil and plant dynamics under irrigation and fertilization is readily available. First numerical results using dynamic programming, treating weather as stochastic uncertainty, are highly promising, setting the stage for this research project.
Keywords
Control Theory, Agriculture, Irrigation, Fertilization, Safety, Stochastic Systems, Reinforcement Learning, Partially Observable Markov Decision Processes
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-19
Applications limited to ETH Zurich
Organization Automatic Control Laboratory
Hosts Schmid Niklas
Topics Mathematical Sciences , Information, Computing and Communication Sciences , Engineering and Technology
Integrated planning and operation of energy systems with seasonal storage
This master thesis develops optimization algorithms to size devices of an energy hub including long-term energy storage, while explicitly considering the operation. The MPC operation must take short-term and seasonal forecasts into account, which represent different time-scales of uncertainty.
Keywords
multi-energy, systems, ehub, planning, design, MPC, bilevel, optimization, forecast, uncertainty, seasonal
Labels
Semester Project , Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-18 , Earliest start: 2026-01-12
Applications limited to Empa , ETH Zurich
Organization Automatic Control Laboratory
Hosts Koepele Cara , Wallington Kevin
Topics Engineering and Technology
Advanced Data Fusion and Signal Interpretation for In-Situ Monitoring and Quality Control in Laser Powder Bed Fusion
This thesis advances in-situ process monitoring in Laser Powder Bed Fusion (PBF-LB/M) by integrating Eddy Current Testing (ECT) and computer vision (CV) for defect detection. The study focuses on signal interpretation across simple to complex geometries, using multi-sensor data fusion to improve defect identification with real world data.
Keywords
Sensor fusion, machine learning, anomly detection, additive manufacturing
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-18 , Earliest start: 2026-02-16 , Latest end: 2026-05-29
Organization Automatic Control Laboratory
Hosts Zakwan Muhammad
Topics Engineering and Technology
Optimization Controller for Gas Compressors Providing Grid Services
This thesis will be conducated in close collaboration with Everllence, a world-leading manufacturer of gas compressors. The aim is to develop an optimization-based controller for gas compressors that provide ancillary services to the power grid. First, the right controller architecture (data-driven, model-based, or hybrid) will be identified. Then, the main objective is to demonstrate the types of grid services that can be offered by compressor plant operators. A cost-benefit assessment of the controller will quantify potential savings and identify key factors influencing the economic feasibility of using compressor applications for grid services.
Keywords
Control, Optimization, Data, Model, Compressor, Power, Energy, Grid, Ancillary, Service
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-18
Organization Automatic Control Laboratory
Hosts Matt Jonas
Topics Engineering and Technology
AI for Grid Stability: Deep Learning the Small-Signal Dynamics of Power Converters
This project focuses on developing and validating a scalable machine learning framework to address the modeling challenges of small-signal stability assessment in grid-connected converters.
Keywords
Data-driven modelling, Machine Learning, Power Converters, Small-signal modelling, Sub-synchronous oscillations, Control, Stability
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-18
Organization Automatic Control Laboratory
Hosts Abdalmoaty Mohamed
Topics Engineering and Technology
Distribution-Free State Estimation for Dynamical Systems via Conformal Prediction
Accurate state estimation in dynamical systems is traditionally achieved with model-based filters and smoothers such as the Kalman filter (KF), particle filters (PF), and Rauch-Tung-Striebel (RTS) smoothers. These methods offer optimal state estimates under specific probabilistic assumptions (e.g. linear Gaussian models) and provide transparency via interpretable models and theoretical guarantees (e.g. MMSE optimality of KF under Gaussian noise). However, real systems often violate these assumptions as noise can be heavy-tailed or unknown and models mismatched leading to overconfident or inaccurate uncertainty estimates. Conformal prediction (CP) provides distribution-free uncertainty quantification with finite-sample coverage guarantees and minimal modeling assumptions. This thesis explores estimation algorithms that combines the transparency and structure of model-based filters/smoothers with CP's distribution-free coverage, studying theory and empirical behavior to obtain the best of both worlds for generic dynamical systems.
Labels
Master Thesis
More information
Published since: 2025-11-18 , Earliest start: 2025-12-01
Organization Automatic Control Laboratory
Hosts Lindemann Lars
Topics Information, Computing and Communication Sciences , Engineering and Technology
Interaction-Aware Control Design with Provable Guarantees in Realistic Robotic Scenarios
This project builds on the guarantees introduced in to design, analyze, and validate an interaction-aware control stack for robotic systems operating among responsive agents. The approach couples an internal trajectory-prediction module that is explicitly updated as policies change with a distribution-free safety layer based on conformal tools that retain finite-sample coverage in the presence of feedback-induced shift and adversarial perturbations. The resulting uncertainty sets are composed with certified planning and safety filters such as control barrier functions and reachability-based shielding. Evaluation encompasses software simulation and on-robot trials in realistic interactive scenarios and compares against confidence-aware and reachability/ORCA-style baselines.
Labels
Master Thesis
More information
Published since: 2025-11-18 , Earliest start: 2025-12-01
Organization Automatic Control Laboratory
Hosts Lindemann Lars
Topics Information, Computing and Communication Sciences , Engineering and Technology
Online Learning-based Control of Indoor CO2 Concentration
We spend more than 90% of our time indoors, hence the air that we breath is at least as important as the food we eat and water we drink. This focus of this project is the control of the indoor air quality (IAQ), in particular the CO2 concentration in the air. Studies have shown, that starting already at concentration levels of 1000ppm in the air, human decision making performance is measurably worse and gets progressively worse as the concentration increases. Modern solutions of keeping this concentration low in, for example, offices, schools or industrial zones, comprise an air damper mounted on the air duct supplying air to the zone and a CO2 sensor that continuously measures the concentration. Air is supplied by fans in the building, and the air damper controls the supply of the air in its corresponding zone. Current control techniques for the air damper are fixed in time and lack the level of sophistication required to adequately monitor the CO2 based on the varying number of occupants in the zone. Moreover, the tuning of the parameters of current approaches is time-consuming and impractical given the high variance in the type and volume of the zones such devices can be deployed at.
Keywords
Online Learning, Adaptive Control, Indoor Air Quality Control, Adaptive Algorithms
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-18 , Earliest start: 2026-01-26 , Latest end: 2026-09-30
Applications limited to ETH Zurich
Organization Automatic Control Laboratory
Hosts Karapetyan Aren
Topics Mathematical Sciences , Information, Computing and Communication Sciences , Engineering and Technology , Behavioural and Cognitive Sciences
Iterative Learning Control for Additive Manufacturing Processes
Additive manufacturing, commonly known as 3D printing, has seen widespread adoption in many engineering fields, including biomedical, aerospace, and automotive applications. Its ability to manufacture complex designs accurately and quickly offers significant advantages in rapid prototyping, customization, and design flexibility for mechanical assembly. As an additive manufacturing technique, fused deposition modeling (FDM) is governed by two key dynamics: motion and extrusion. In FDM, a triple axis system controls the movement of the extrusion head, which melts and deposits plastic filament onto the printing bed. However, these dynamics are inherently coupled, whereby the extrusion width is directly affected by the motion of the system. Given that the printing motion is highly repetitive, it is advantageous to leverage this behavior when designing a control system.
Keywords
Motion control, adaptive control, learning control, robotics
Labels
Semester Project , Master Thesis
PLEASE LOG IN TO SEE DESCRIPTION
More information
Published since: 2025-11-18 , Earliest start: 2025-11-25
Applications limited to ETH Zurich
Organization Automatic Control Laboratory
Hosts Hoteit Rawan , Zuliani Riccardo , Balta Efe
Topics Engineering and Technology
Safe Manipulation in Complex Environments
Dynamical-system approaches for robotic control reshape velocity fields around obstacles for im- pressive reactive avoidance. Yet, they can still encounter unwanted local minima and saddle points. Research on Control Lyapunov Functions (CLF) and Control Barrier Functions (CBF) has estab- lished conditions under which such equilibria arise and has provided controller-synthesis methods to avoid them. By incorporating these safety and stability frameworks into DS approaches we hope to introduce new solutions for real-time task execution in cluttered, real-world settings.
Keywords
Robotics, Collision Avoidance
Labels
ETH Zurich (ETHZ)
Description
Goal
Contact Details
More information
Published since: 2025-11-17 , Earliest start: 2026-01-01 , Latest end: 2026-08-01
Organization ETH Zurich
Hosts Stocco Paula , Wang Han
Topics Mathematical Sciences , Information, Computing and Communication Sciences
Uncertainty Modeling for Robotic Trajectory Optimization
Machine learning is increasing integrated into control systems as accurate and fast system models capable of representing nonlinearities. Gaussian Processes (GPs) are stochastic processes that can be used as a surrogate model for the system. They provide not only a prediction but a quantification of uncertainty, making them especially appealing for safety-critical systems. Despite the success of GPs in modeling system dynamics for data-efficient model-based reinforcement learning and safe Model Predictive Control (MPC), little work has studied the accuracy of GP derivatives, or how those errors propagate through optimization-based controllers. This project will contribute a theoretically grounded method for incorporating gradient uncertainty into control optimization, aiming to improve both safety and convergence time.
Keywords
Uncertainty Quantification, Trajectory Following, Robotics, Machine learning, Optimization
Labels
ETH Zurich (ETHZ)
Description
Goal
Contact Details
More information
Published since: 2025-11-17 , Earliest start: 2026-01-01 , Latest end: 2026-08-01
Organization Automatic Control Laboratory
Hosts Stocco Paula , Zuliani Riccardo
Topics Information, Computing and Communication Sciences
Verifiable Control Design with Data-Driven Spectral Submanifolds
Low-dimensional latent space representations of dynamical systems provide a powerful tool for scalable control design. Over the last years, learning-based approaches for constructing latent space representations have gained popularity, often utilizing variational autoencoders. While these approaches have shown great empirical success, they typically lack formal control guarantees as needed in safety-critical applications. In recent work, we presented a framework for designing controllers in such learned latent spaces that can provably guarantee stability and safety for the original system by exploiting approximate conjugacy between the latent and full dynamics. However, large approximate conjugacy can result in overly conservative and poorly performing control actions. On the other hand, spectral submanifolds were studied in as attracting invariant manifolds. Spectral submanifolds guarantee the existence of a low-dimensional latent space representation that, if restricted to the manifold, have the same behavior as the original system. Spectral submanifolds are constructed from the linear components of a dynamical system and have zero conjugacy error, even though their data-driven construction is approximate in nature. This thesis will investigate how to integrate the framework of spectral submanifolds into the latent space control design framework from \cite{lutkus2025latent} to achieve non-conservative control behavior while retaining stability and safety guarantees. The project combines insights from dynamical systems, control theory, and data-driven reduced order models.
Labels
Master Thesis
More information
Published since: 2025-11-17 , Earliest start: 2025-12-01
Organization Automatic Control Laboratory
Hosts Lindemann Lars
Topics Engineering and Technology
Verifiable Latent Space Control Design Beyond Stability and Forward Invariance
Low-dimensional latent space representations of dynamical systems provide a powerful tool for scalable control design. Over the last years, data-driven approaches for constructing latent space representations have gained popularity and shown great empirical success. While these methods are promising, they typically lack formal control guarantees as needed in safety-critical applications. In recent work, we provided a theoretical framework for designing controllers in such learned latent spaces that can provably guarantee stability and safety for the original system by exploiting approximate conjugacy between the latent and full dynamics. Yet, these guarantees are largely limited to using notions of Lyapunov and barrier functions, ensuring stability and forward invariance only. Modern autonomous control systems, however, must often satisfy richer temporal or logical specifications that involve deadlines, sequencing, and reactive behavior. This thesis will investigate how to extend our latent space control design framework beyond stability and forward invariance, towards achieving verifiable temporal and logic-based system behavior. The project combines insights from representation learning, formal methods, and control theory, aiming to unify latent space learning with verifiable control design.
Labels
Master Thesis
More information
Published since: 2025-11-17 , Earliest start: 2025-12-01
Organization Automatic Control Laboratory
Hosts Lindemann Lars
Topics Information, Computing and Communication Sciences , Engineering and Technology
Is oversizing a mistake? Optimal heat pump capacities considering flexibility potential
This semester project explores the trade-offs between heat pump oversizing and flexibility potential through simulation. Ideally, an optimization algorithm considering the heat pump operation is developed and adapted to size the heat pump given flexibility constraints.
Keywords
heat, pump, size, capacity, flexibility, control, optimization, simulation
Labels
Semester Project
PLEASE LOG IN TO SEE DESCRIPTION
More information
Published since: 2025-11-17 , Earliest start: 2026-01-12
Organization Automatic Control Laboratory
Hosts Koepele Cara
Topics Engineering and Technology
Behind Ride Fares: Dynamic Pricing for Mobility Systems
Imagine you are asked by Uber to set their base prices and cross-region prices for daily travel requests in Zürich. You would like to maximize the potential profit of Uber. Fortunately, you have rough ideas on the demand levels of Zürich and how they are affected by your chosen prices. Nonetheless, such random demands fluctuate every day, and they also drift with the conditions of other mobility service providers. How will you determine ideal Uber prices? The above problem is an instance of stochastic optimization with decision dependence, where a decision maker interacts with a changing data distribution. We will develop an online algorithm based on daily samples and real-time interactions to navigate through the space of pricing vectors, thereby maximizing expected profits. We will leverage a dedicated agent-based simulator to demonstrate the effectiveness of the above strategy and compare it against benchmark schemes.
Keywords
Mobility systems, dynamic pricing, online optimization
Labels
Semester Project , Bachelor Thesis , Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-17 , Earliest start: 2026-01-01 , Latest end: 2026-08-31
Organization Automatic Control Laboratory
Hosts He Zhiyu
Topics Mathematical Sciences , Information, Computing and Communication Sciences , Engineering and Technology
Self-supervised parallel manipulation learning via disagreement
Imitation Learning has proven able to replicate the fine dexterous manipulation movements by replicating human demonstrations. However, collecting such demonstrations is difficult to scale due to the need for a human expert in the data collection loop. A solution to this problem can be found in self-supervised exploration via disagreement. Exploration by disagreement takes inspiration from classical active learning literature to devise a self-supervised procedure able to efficiently explore the model space.
Keywords
Robotics, Control, Dexterous Manipulation
Labels
Master Thesis , ETH Zurich (ETHZ)
Description
Goal
Contact Details
More information
Published since: 2025-11-15 , Earliest start: 2026-01-01 , Latest end: 2026-07-01
Organization Automatic Control Laboratory
Hosts Rimoldi Alessio
Topics Information, Computing and Communication Sciences
Deep reinforcement learning for job shop scheduling
Production scheduling is a critical event in the production management of smart manufacturing systems. Many job shop scheduling (JSP) problems are known as NP-hard. Meanwhile, in the real production plants, not only is the objective of scheduling usually multiple (e.g., makespan, tardiness, machine utilization), but also the environment is mostly dynamic with unexpected machine breakdowns, order arrivals and cancellations, etc. Because the JSP can be modelled as a Markov Decision Process (MDP), using deep reinforcement learning techniques for scheduling has gained significant attention in this domain. In this project, a deep reinforcement learning environment for JSP will be developed and enhanced to tackle the aforementioned challenges.
Keywords
Deep reinforcement learning, Job shop scheduling, Operational research, Multi-objective optimization
Labels
Semester Project , Master Thesis
Description
Contact Details
More information
Published since: 2025-11-15 , Earliest start: 2026-01-01 , Latest end: 2026-09-01
Applications limited to ETH Zurich
Organization Automatic Control Laboratory
Hosts Tang Yu, Mr.
Topics Engineering and Technology
Temporally Robust Controller Synthesis for Time-Critical Systems
The reliability of autonomous control systems depends on their robustness, i.e., their ability to meet their objectives under uncertainty. Emerging applications in autonomy are increasingly time-critical: failing to meet temporal requirements can severely compromise safety and performance, as in multi-robot disaster response, fleets of autonomous taxis, and automated airport ground control. Yet today’s technology still struggles with unpredictable delays and coordination failures, limiting their reliability in real-world settings. Existing research has focused on spatial robustness, ensuring that spatial objectives - such as collision avoidance of a robot - are met despite modeling errors and disturbances. However, time-critical systems also require temporal robustness, the ability to meet time-critical objectives - such as deadlines, sequencing, and periodic tasks - under uncertainty. A major challenge arises from variations in computation and actuation times, and particularly from complex timing uncertainties caused by human interaction, unpredictable sensing failures, and compute-intensive perception. This thesis will study novel ways to enable temporal robustness of autonomous control systems.
Labels
Master Thesis
More information
Published since: 2025-11-14 , Earliest start: 2025-12-01
Organization Automatic Control Laboratory
Hosts Lindemann Lars
Topics Information, Computing and Communication Sciences , Engineering and Technology
Feedback Control for Learning in Optimization and Game Theory
How can we design learning dynamics that are fast, robust, and provably correct when many decision-makers interact under shared constraints? This project investigates this question by leveraging a novel control-theoretic framework for equality-constrained optimization to design and analyze dynamics that compute solutions to Generalized Nash Equilibrium problems in both static and time-varying settings. Using contraction theory as a scalable framework for robust stability analysis, the goal is to derive verifiable conditions for global exponential convergence and robustness to disturbances, leading to algorithms with provable performance guarantees.
Keywords
Generalized Nash Equilibrium, Contraction Theory, Primal--Dual Dynamics, Feedback Control, Time-Varying Games
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-14 , Earliest start: 2025-12-01
Applications limited to ETH Zurich , University of Zurich
Organization Automatic Control Laboratory
Hosts Centorrino Veronica
Topics Mathematical Sciences
Grid-Connected Electrolyzer for Dynamic Ancillary Services Provision
This master thesis explores how hydrogen electrolyzers can contribute to power system stability through the provision of dynamic ancillary services while producing renewable hydrogen. In collaboration with H2 Energy, the project combines system modeling, control design, and experimental validation to investigate how grid-connected electrolyzers can deliver a range of grid services. The work aims to establish a comprehensive understanding of the technical feasibility, control strategies, and operational trade-offs associated with such multi-service operation without compromising electrolyzer health, efficiency, or hydrogen yield.
Keywords
electrolyzer, dynamic grid services, power electronics control design, hydrogen, experimental validation
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-14 , Earliest start: 2026-02-01
Organization Automatic Control Laboratory
Hosts Häberle Verena
Topics Engineering and Technology
Fair Demand-Side Flexibility Allocation with Karma
The rapid integration of distributed renewable energy sources into electric power grids introduces the challenges of balancing intermittent power production and guaranteeing grid stability. Demand-side energy flexibility, which involves the end-user shifting or adjusting energy consumption in line with the supply and capacity of the electricity grid, is receiving increasing attention as an important approach towards tackling these challenges. Buildings are significant energy consumers and hence present a promising source of flexibility. Leveraging the thermal energy storage capabilities of buildings allows to intentionally modify their energy consumption to support grid needs without compromising occupants’ thermal comfort. Previous work has focused on quantifying energy flexibility envelopes or designing temperature control strategies which are important to meet flexibility objectives. However, in practice, Distribution System Operators (DSOs) often prefer to provide flexibility targets to groups of buildings rather than to individual units. This leads to the question of how to fairly and efficiently distribute flexibility targets to individual households within a designated group. This project aims to develop a karma mechanism for this flexibility allocation problem. Karma economies are a promising recently developed non-monetary solution to the allocation of shared resources. These economies leverage the fact that shared resources, such as flexibility requests, do not occur once but are repeated frequently. It is thus envisioned that karma will enable consumers to express when they have high urgency to avoid shifting their consumption, meanwhile guaranteeing that everyone contributes fairly to providing flexibility over time.
Labels
Master Thesis
Description
Contact Details
More information
Published since: 2025-11-12 , Earliest start: 2025-12-01
Organization Automatic Control Laboratory
Hosts Montazeri Mina , Elokda Ezzat , Nortmann Benita
Topics Engineering and Technology
The “Give-and-Take” in Recommender Systems: A Karma Economy for Fair Exploration
On platforms like Spotify or YouTube, new content is constantly uploaded. They must be explored by different users before the recommender system understands the quality and best audience. However, users usually prefer high-quality content. Showing too much new content can negatively affect their experience. This creates an important challenge for recommender systems: Who should receive exploratory recommendations (new content) and who should receive exploitative recommendations (high-quality content). To do this fairly, we propose a Karma Economy to manage it, where users can 1. spend Karma to get top-quality recommendations (exploit), or 2. earn Karma by helping evaluate new or uncertain content (explore).
Keywords
Recommender Systems, Multi-Agent Systems, Fairness, Karma Economy
Labels
Semester Project , Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-12 , Earliest start: 2025-12-01 , Latest end: 2026-09-30
Organization Automatic Control Laboratory
Hosts Zhang Kai
Topics Mathematical Sciences , Information, Computing and Communication Sciences , Economics
Online Feedback Optimization for Real Time Power Grid Control
Online Feedback Optimization (OFO) allows to drive a dynamical system to a steady state that satisfies given optimality criteria, such as efficiency and satisfactions of constraints. In the past years, we had great success in applying OFO to the real-time control of power grids, both for energy distribution and transmission. With RTE France we simulated how to solve grid congestion when there is an excess of wind generation. We deployed an OFO algorithm on the Swiss grid to control the voltage and the reactive power of a distribution grid. And we won the Watt D'Or award by the Swiss Federal Office of Energy!
Keywords
smart grid, optimization, control of power grids, feedback optimization, energy
Labels
Semester Project , Bachelor Thesis , Master Thesis
Description
More information
Published since: 2025-11-11 , Earliest start: 2026-01-01
Organization Automatic Control Laboratory
Hosts Bolognani Saverio
Topics Information, Computing and Communication Sciences
Games in Motion: Learning Equilibria in Metric Spaces
Imagine a strategic competition among multiple parties for the attention of the general public with complex interactions. These can be a Democrat and a Republican competing for votes across a large population, or Pepsi and Cola battling for market shares in a vast region. What are the possible outcomes? How can one gain an edge compared to the opponent? These interactions can be characterized as equilibrium-seeking problems in metric probability spaces, featuring strategic decision-making under evolving distribution dynamics. We will bridge insights from game theory, dynamical systems, complex networks, and optimal transport to shed light on solution concepts, algorithmic pipelines, and performance guarantees in such non-stationary environments.
Keywords
Game theory, dynamics, decision dependence, metric probability spaces.
Labels
Semester Project , Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-11-10 , Earliest start: 2025-12-01 , Latest end: 2026-08-31
Organization Automatic Control Laboratory
Hosts He Zhiyu
Topics Mathematical Sciences , Information, Computing and Communication Sciences , Engineering and Technology
A Game Plan for Electricity: Optimal Incentive Design for Procuring Voltage Control in Transmission Grids
The reliable operation of modern power systems increasingly relies on flexible and distributed resources such as renewable generators, battery storage, and demand-side participants. These resources can provide essential control services such as frequency and voltage support. As many of these resources participate voluntarily, system operators must design incentive mechanisms that encourage participation while ensuring safe and stable grid operation. Existing approaches for procuring and coordinating such services often lack formal guarantees regarding stability, robustness, and economic efficiency, especially under uncertainty in system dynamics, participant behavior, or external disturbances. This creates an opportunity to develop new incentive and control strategies that combine theoretical rigor with practical applicability.
Keywords
Incentive Design, Game Theory, Power Grid, Energy Systems, Swissgrid, Voltage, Control, Optimization
Labels
Semester Project , Bachelor Thesis , Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-10-29
Applications limited to ETH Zurich
Organization Automatic Control Laboratory
Hosts Matt Jonas
Topics Mathematical Sciences , Engineering and Technology
Reading the Game: Predicting the Behavior of Participants in Swissgrid’s Voltage Control Program via Inverse Optimization and Inverse Game Theory
The secure and efficient operation of the Swiss transmission grid relies on the cooperation of connected agents, such as power plants and distribution system operators (DSOs), which can provide voltage support through reactive power control. To promote this behavior, Swissgrid offers financial incentives via its voltage support program. However, operational data shows that participants react very heterogeneously to these incentives. The reasons for this diversity are unclear, as the agents’ cost structures, technical limits, and strategic motivations are not directly observable. Understanding this behavioral diversity is key to designing more effective and equitable incentive mechanisms for future power systems.
Keywords
Inverse, Optimization, Game Theory, Control, Incentive Design, Learning, Stackelberg, Leader, Follower, Bilevel, Swissgrid, Power Grid, Energy Systems
Labels
Semester Project , Bachelor Thesis , Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-10-29
Applications limited to ETH Zurich
Organization Automatic Control Laboratory
Hosts Matt Jonas
Topics Mathematical Sciences , Engineering and Technology
Enhancing Stability of Large-Scale Power Systems via Learning Dissipativity
Modern power systems are nonlinear, complex, and interconnected with numerous heterogeneous components, causing significant challenges to system stability. Control theory techniques that provide stability guarantees typically rely on a simplified model and do not capture the nonlinear behavior of the dynamics, motivating a deep-learning-based approach. However, naive deep-learning-based approaches generally suffer from the scale of dimensionality, especially in the context of large-scale power systems. Therefore, this project aims to develop a deep learning-based controller in a decentralized fashion based on dissipativity theory, in order to ensure the global stability of the system in a scalable fashion.
Keywords
Deep Learning, Interconnected System, Power System Transient Stability
Labels
Master Thesis , ETH Zurich (ETHZ)
PLEASE LOG IN TO SEE DESCRIPTION
More information
Published since: 2025-10-24 , Earliest start: 2025-10-27
Organization Automatic Control Laboratory
Hosts Wang Han , Nakano Taiki
Topics Engineering and Technology
Small-Signal Data-Driven Modeling of Power Converters Under Unbalanced Conditions
This project focuses on developing and validating data-driven small-signal models of grid-connected converters under unbalanced and harmonic conditions. The identified models are crucial to understand sub-synchronous oscillations caused by converter-grid interactions.
Keywords
Data-driven modelling, System Identification, Power Converters, Small-signal modelling, Sub-synchronous oscillations, Control, Stability, Unbalanced, Harmonics
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-10-22
Organization Automatic Control Laboratory
Hosts Abdalmoaty Mohamed
Topics Engineering and Technology
Multi-Agent Data-Driven Control for Power Oscillation Damping
This project investigates multi-agent data-driven control as a novel approach to damping oscillations in converter-dominated power systems, where classical model-based methods are increasingly unreliable. The study will compare a centralized benchmark with decentralized schemes, assessing how locally updated controllers can collectively achieve coordinated, system-wide stability.
Keywords
data-driven control, multi-agent system, dynamic power system modelling
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-10-02 , Earliest start: 2026-01-01
Organization Automatic Control Laboratory
Hosts Häberle Verena
Topics Engineering and Technology
Multi-Agent Grid Impedance Identification in Three-Phase Power Systems
This project investigates multi-agent grid impedance identification in three-phase power systems. It will develop and compare two approaches: a global multi-port identification framework and a locally simultaneous multi-agent single-port identification framework. The study will evaluate and compare the accuracy of the two approaches and explore potential downstream applications in stability analysis and control design.
Keywords
grid impedance, system identification, multi-agent systems, three-phase power system dynamic modelling
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-10-01 , Earliest start: 2026-01-01
Organization Automatic Control Laboratory
Hosts Häberle Verena
Topics Engineering and Technology
Optimal Excitation for Grid Impedance Estimation
This project aims to develop optimal excitation schemes for impedance estimation of grid/grid-connected converters.
Keywords
Impedance estimation; grid-connected converters; optimal excitation; experiment design; system identification
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-09-22 , Earliest start: 2025-09-21
Applications limited to ETH Zurich
Organization Automatic Control Laboratory
Hosts He Xiuqiang , Abdalmoaty Mohamed
Topics Engineering and Technology
Adaptive control via reinforcement learning: stability, optimality, and robustness
This project explores reinforcement learning (RL) for adaptive control of linear time-invariant systems, with a focus on achieving stability, optimality, and robustness. While RL-based adaptive control methods are gaining popularity, most lack rigorous stability guarantees, especially when applied to the linear quadratic regulator (LQR) problem. Building on recent advances in sequential stability analysis, the project aims to develop RL algorithms that ensure closed-loop stability and convergence to the optimal LQR policy. Theoretical insights will be validated through simulations on representative control systems.
Keywords
data-driven control, adaptive control, reinforcement learning, linear time-invariant system
Labels
Master Thesis , ETH Zurich (ETHZ)
Description
Goal
Contact Details
More information
Published since: 2025-09-09 , Earliest start: 2025-09-07 , Latest end: 2026-07-01
Organization Automatic Control Laboratory
Hosts Bartos Marcell , Zhao Feiran
Topics Mathematical Sciences , Information, Computing and Communication Sciences
Safe and Reliable Model Predictive Control using Differentiable Optimization
Safety violations in control systems can lead to catastrophic outcomes, from autonomous vehicle crashes to power grid failures. While Model Predictive Control (MPC) offers powerful safety mechanisms through constraint enforcement, a critical dilemma emerges: improved controller performance often comes at the expense of safety margins. Traditional tuning approaches that prioritize performance metrics may inadvertently compromise safety guarantees. This project addresses this fundamental challenge by developing a tuning framework that enhances MPC performance while providing anytime safety guarantees—ensuring the system remains safe even during ongoing optimization. The approach offers a principled solution for deploying high-performance, safety-critical control in autonomous systems, robotics, and industrial processes.
Keywords
Model Predictive Control, Learning-based Control, Differentiable Optimization
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-09-02 , Earliest start: 2025-07-01
Organization Automatic Control Laboratory
Hosts Zuliani Riccardo
Topics Engineering and Technology
High-performance Model Predictive Control for Autonomous Driving via Policy Optimization, in Collaboration with AMZ
Autonomous driving demands controllers that combine high performance, safety, and real-time feasibility. While Model Predictive Control (MPC) meets these requirements in principle, practical implementations often rely on simplified models and short horizons, leading to suboptimal performance. This project aims to develop a hyperparameter tuning scheme that optimizes MPC parameters to recover near-optimal behavior without increasing model complexity or prediction horizon. The method will be validated in simulation and, for interested students, tested on an autonomous racing car, aiming to reduce lap times while ensuring compliance with safety constraints such as tire friction and lane boundaries.
Keywords
Autonomous Driving, Model Predictive Control, Learning-based Control
Labels
Semester Project , Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-09-02 , Earliest start: 2025-09-14
Organization Automatic Control Laboratory
Hosts Hoteit Rawan , Zuliani Riccardo
Topics Engineering and Technology
Semantic Segmentation for Volume Estimation
This thesis investigates the use of vision foundation models for semantic segmentation within 3D reconstruction pipelines to improve volume estimation in industrial settings. Using multi-view datasets from Tinamu Labs, the work focuses on segmenting stockpiles, static warehouse structures, and occluding objects. The approach combines geometric information with segmentation models and addresses occluded or missing regions through automatic detection and infill. The outcome supports more accurate and robust volume estimation, contributing to automated inventory management. The project is conducted in collaboration with Tinamu Labs and validated on their robotic systems.
Keywords
Data analysis, machine learning, semantic segmentation
Labels
Master Thesis
Description
Goal
Contact Details
More information
Published since: 2025-09-01 , Earliest start: 2025-09-15 , Latest end: 2025-12-19
Organization Automatic Control Laboratory
Hosts Zakwan Muhammad
Topics Engineering and Technology
Dynamic Kicking Skills for Humanoid Robots
Powerful yet accurate kicking is an essential skill for humanoid robots, especially in the context of robotic soccer competitions like RoboCup. Developing effective kicking skills for humanoid robots is a complex task that involves a combination of mechanical design, advanced control algorithms, and sensor integration. This project aims to develop a robust control system for a NAO robot to successfully kick a rolling ball into a designated goal area while dynamically tracking the ball position.
Keywords
Motion control, Reference Tracking, Robotics
Labels
Semester Project
Description
Goal
Contact Details
More information
Published since: 2025-08-13 , Earliest start: 2025-09-16 , Latest end: 2026-03-31
Applications limited to ETH Zurich
Organization Automatic Control Laboratory
Hosts Hoteit Rawan , Stocco Paula
Topics Engineering and Technology
Modeling a tri-winged airborne wind turbine, a data-driven approach
Wind energy is key to the green transition, but traditional turbines are costly and long to build. Airborne Wind Energy (AWE) offers a lighter, cheaper alternative by using tethered wings to harness stronger winds at higher altitudes. We have developed a novel AWE system with three interconnected wings orbiting each other—early results show it’s not only easier and cheaper to build but also offers better control than current AWE designs. We need an accurate and robust dynamic model of the system. In this project, the student will use system identification techniques to derive models of the three-wing AWE system. You will work with both simulation data and measurements from a small-scale prototype, with the goal of delivering a validated identification pipeline that will be tested on a larger prototype at the end of the project.
Labels
Master Thesis , ETH Zurich (ETHZ)
Description
Contact Details
More information
Published since: 2025-07-28 , Earliest start: 2025-09-07 , Latest end: 2026-04-30
Organization Automatic Control Laboratory
Hosts Brouillon Jean-Sébastien
Topics Engineering and Technology
Modeling a tri-winged airborne wind turbine, first principles
Wind energy is key to the green transition, but traditional turbines are costly and long to build. Airborne Wind Energy (AWE) offers a lighter, cheaper alternative by using tethered wings to harness stronger winds at higher altitudes. We have developed a novel AWE system with three interconnected wings orbiting each other—early results show it’s not only easier and cheaper to build but also offers better control than current AWE designs. Before control strategies, safety validations, and certifications can be addressed, we need an accurate and robust dynamic model of the system. In this project, the student will use theoretical first principles from fluid dynamics to derive good model candidates, in increasing levels of detail and complexity. Initial values of the model parameters should be provided based on airfoil data and/or computational fluid dynamics.
Keywords
Green energy, aeronautics, dynamical models
Labels
Master Thesis , ETH Zurich (ETHZ)
Description
Contact Details
More information
Published since: 2025-07-28 , Earliest start: 2025-09-07 , Latest end: 2026-04-30
Organization Automatic Control Laboratory
Hosts Brouillon Jean-Sébastien
Topics Engineering and Technology
Disturbance rejection for a tri-winged airborne wind turbine
Wind energy is key to the green transition, but traditional turbines are costly and long to build. Airborne Wind Energy (AWE) offers a lighter, cheaper alternative by using tethered wings to harness stronger winds at higher altitudes. We have developed a novel AWE system with three interconnected wings orbiting each other—early results show it’s not only easier and cheaper to build but also offers better control than current AWE designs. While our system is passively stable, its sensitivity to disturbances from wind gusts and other sources must be quantified to obtain the required safety margins. Moreover, several active control architectures will be explored to reduce this sensitivity as much as possible. You will work with both a comprehensive simulation framework and a small-scale prototype, with the goal of delivering a disturbance sensitivity analysis that will be tested in-field on a larger prototype at the end of the project. This thesis is part of the foundational work for a startup aiming to bring this innovative concept into real-world applications.
Keywords
Green energy, aeronautics, control systems
Labels
Master Thesis , ETH Zurich (ETHZ)
Description
Contact Details
More information
Published since: 2025-07-28 , Earliest start: 2025-09-07 , Latest end: 2026-04-30
Organization Automatic Control Laboratory
Hosts Brouillon Jean-Sébastien
Topics Engineering and Technology
Regulatory framework for airborne wind energy systems
Wind energy is key to the green transition, but traditional turbines are costly and long to build. Airborne Wind Energy (AWE) offers a lighter, cheaper alternative by using tethered wings to harness stronger winds at higher altitudes. We have developed a novel AWE system with three interconnected wings orbiting each other—early results show it’s not only easier and cheaper to build but also offers better control than current AWE designs. AWE systems are flying objects, which are strictly regulated. Although our breakthrough can allow for lighter and safer wings, a close contact with authorities is required to avoid unnecessary risks later on.
Keywords
Green energy, aeronautics, safety, regulations
Labels
Semester Project , Master Thesis , ETH Zurich (ETHZ)
Description
Contact Details
More information
Published since: 2025-07-28 , Earliest start: 2025-08-05 , Latest end: 2026-07-31
Organization Automatic Control Laboratory
Hosts Brouillon Jean-Sébastien
Topics Law, Justice and Law Enforcement , Engineering and Technology
Contact
ETH Zurich
Automatic Control Laboratory
Physikstrasse 3
8092 Zurich
Switzerland