Student Projects

The goal of this project is to expand the data-driven controller synthesis framework developed at the lab to address performance requirements that are not adequately captured by the objectives currently available. At present, the framework supports (H_2) and (H_\infty) objectives and constraints, as well as robustness specifications based on Integral Quadratic Constraints (IQCs). However, practical requirements such as actuator saturation limits may require alternative performance measures and synthesis formulations.

The project will focus on the following components:

* Theoretical: Investigate alternative control objectives, such as (H_1) performance and regret-based criteria, and develop corresponding data-driven controller synthesis methods.

* Software: Implement the developed algorithms in the lab’s `datadriven` software library, with an emphasis on numerical reliability and integration with the existing framework.

* Experimental: Validate the proposed methods on real-world experimental systems available at the lab, with particular attention to their ability to handle saturation-related constraints and performance requirements.

Professor(s)

Alireza Karimi, Vaibhav Gupta

Administration

Barbara Marie-Louise Frédérique Schenkel

Site

ddmac.epfl.ch, la.epfl.ch

The EPFL babyfoot is under continuous improvement.

While the babyfoot can easily intercept the ball and kick toward the opponent goal with success, it can also juggle and shoot if the ball is still.

The difficulty is to capture the ball, make a pass and chain precise actions. This project aims at improving the ball handling while it is moving to permit capture, pass and juggling, and then shoot toward opponents goal.

Students suggested improvements are also welcome.

The babyfoot is programmed using LabVIEW, knowledge of this language is prerequisite for strategy related projects.

Professor(s)

Christophe Salzmann (Laboratoire d’automatique 3)

Site

https://www.epfl.ch/labs/la/pi/babyfoot/

A widespread challenge in industrial automation is that complex systems are governed by PID controllers that, over time, become poorly tuned or drift from their original settings, silently degrading performance and product quality. This is a core problem in process industries (chemical, thermal, manufacturing) and a key bottleneck on the path to Industry 4.0 and autonomous plants. The student will first develop a model of an industrial thermal process – such as an industrial oven, dryer, or reactor – and implement a digital twin simulator with a PID controller in the loop. In a second phase, state-of-the-art neural network optimal control methods will be applied to go beyond what a detuned PID can achieve, bridging classical control theory with modern learning-based control.

Objectives

– Review the literature and identify a suitable mathematical model for a chosen type of systems.

– Implement a simulator (digital twin) with a baseline PID controller in the loop.

– Study and apply neural network optimal control methods to improve closed-loop performance.

Methodology

– Conduct a concise literature survey on the chosen process and PID-based industrial control.

– Implement and validate the simulator in Python.

– Benchmark PID performance and characterize its limitations under detuned/drifted conditions.

– Design and train neural network controllers to improve upon baseline PID performance.

Required Skills & Tools

– Control Theory

– Machine Learning (PyTorch)

– Python & MATLAB

Keywords: Digital Twins · Neural Optimal Control · Industry 4.0 · Learning-Based Control · Process Automation · PID Retuning · Sim-to-Real

Comment

If you are interested, please send your transcript (BSc and MSc) to the following email: [email protected] [email protected]

Professor(s)

Giancarlo Ferrari Trecate

Administration

Barbara Marie-Louise Frédérique Schenkel

Context.

Modern control systems operate in environments where disturbances are not stationary but depend on observable contextual variables such as weather, time of day, or operating mode. Classical robust control guards against all disturbances within a fixed, context-free uncertainty set, leading to overly conservative policies. Data-driven distributionally robust control reduces conservatism by learning the uncertainty structure from historical data, but still ignores available contextual information. Contextual robust optimization has recently shown that conditioning the uncertainty set on observable context yields smaller, yet statistically valid, sets that bypass some of the overly conservative nature of robust control. Bringing this paradigm into closed-loop control, with provable coverage guarantees and no distributional assumptions, is an open and timely research challenge.

Main objective. Develop a contextual data-driven robust control framework that constructs context-dependent uncertainty sets from historical data using conformal prediction while guaranteeing marginal coverage of the true disturbance with high probability.

Subobjective 1 (Theory). Model the contextual data-driven robust control problem theoretically and relate to the current literature on robust control.

Subobjective 2 (Experiments). Validate the framework on benchmark control tasks.

Profile of the Candidate:

Strong background in control theory; Solid knowledge of optimization (robust and convex); Familiarity with machine learning theory; Good experience with Python; Desire to participate in a research project with a likelihood of publishing.

Comment

If interested, please reach out by email to Julien Pallage and Sabri El Amrani at [email protected] and [email protected].

Professor(s)

Giancarlo Ferrari Trecate, Julien Pallage

Administration

Barbara Marie-Louise Frédérique Schenkel

ROJECT SUMMARY

Adaptive optics (AO) systems are critical in overcoming atmospheric distortions in ground-based telescopes, enabling sharper and more detailed astronomical observations. This project focuses on designing and synthesising a robust, real-time controller for an AO system. By leveraging advanced control theory, the project aims to significantly enhance the performance of AO systems, ensuring reliable correction of wavefront distortions and improving the quality of astronomical images.

BACKGROUND AND MOTIVATION

Telescopes observing through Earth’s atmosphere suffer from distortions caused by turbulent air layers, which degrade image resolution. Adaptive optics counteract this by using deformable mirrors and wavefront sensors to correct for distortions in real-time.

Current AO systems, while effective, face limitations in:

* Handling dynamic, unknown atmospheric conditions.

* Scalability for large telescopes and next-generation systems.

By synthesising a controller tailored to these challenges, this project seeks to push the boundaries of AO performance, paving the way for discoveries in astronomy and astrophysics.

OBJECTIVES

1. Design and develop a robust controller that addresses dynamic atmospheric variations.

2. Validate performance through simulation (and hardware) implementation using an AO testbed.

3. Ensure scalability and adaptability of the controller for different telescopes and operational conditions.

REQUIREMENTS

The project demands a solid academic foundation in control courses, particularly in ‘Advanced Control Systems.’ Proficiency in the frequency domain approach and robust control techniques is especially critical.

Professor(s)

Alireza Karimi, Vaibhav Gupta

Administration

Barbara Marie-Louise Frédérique Schenkel

External

Department of Astronomy, UNIGE, Isaac Dinis, [email protected]

Site

ddmac.epfl.ch, la.epfl.ch

The goal of this project is to expand the data-driven controller synthesis framework developed at the lab to address performance requirements that are not adequately captured by the objectives currently available. At present, the framework supports (H_2) and (H_\infty) objectives and constraints, as well as robustness specifications based on Integral Quadratic Constraints (IQCs).

The project will focus on the following components:

* Theoretical: Investigate alternative control objectives, in particular Distributionally Robust Controls (DRO), and develop corresponding data-driven controller synthesis methods.

* Software: Implement the developed algorithms in the lab’s `datadriven` software library, with an emphasis on numerical reliability and integration with the existing framework.

* Experimental: Validate the proposed methods on real-world experimental systems available at the lab, with particular attention to their ability to handle saturation-related constraints and performance requirements.

Professor(s)

Alireza Karimi, Vaibhav Gupta

Administration

Barbara Marie-Louise Frédérique Schenkel

Site

ddmac.epfl.ch, la.epfl.ch

This project explores the implementation of a partial feedback linearization algorithm for the swing up control of an inverted pendulum. The algorithm will also be extended to the data-driven paradigm. The designed controller will be implemented on the hardware setup using LabVIEW.

Professor(s)

Alireza Karimi, Vishnu Varadan

Administration

Barbara Marie-Louise Frédérique Schenkel

Site

https://www.epfl.ch/labs/la/, https://www.epfl.ch/labs/ddmac/

The goal of the project is to improve the performance of three non-holonomic bumper cars, operating within a bounded 7×10 arena, that have to reach an assigned target position while achieving a prescribed final orientation. The central difficulty is that the non-holonomic nature of the vehicle dynamics precludes stabilization to a point by simple smooth static state-feedback controllers (a consequence of Brockett’s necessary condition) and therefore we need a more expressive control architecture. The main idea is to use reference Performance Boosting (rPB), which augments a base controller with a learned performance-boosting component while preserving closed-loop stability guarantees. The controller is designed and trained on top of a data-driven MLP model of the bumper cars that was obtained from real data.

Professor(s)

Giancarlo Ferrari Trecate, Simone Baratto

Administration

Barbara Marie-Louise Frédérique Schenkel

**Project Overview**

The project explores meta‑learning methods to develop neural‑network controllers that adapt rapidly to changing disturbance profiles, such as varying wind conditions. By leveraging prior experience across tasks, the controller should generalize efficiently to new scenarios with minimal additional data.

**Objectives**

– Review state-of-the-art meta‑learning algorithms (e.g., MAML) and their suitability for control.

– Analyze neural‑network control architectures and identify key design choices for disturbance rejection.

– Implement selected meta‑learning algorithms in Python (using PyTorch) within a simulated quadcopter environment.

– Evaluate adaptation speed and robustness under diverse wind profiles and quantify improvements over baseline controllers.

**Methodology**

– Conduct a concise literature survey on meta‑learning and neural control.

– Select and configure a quadcopter simulator for disturbance testing.

– Train meta‑learned controllers and compare against standard training (from scratch).

**Required Skills & Tools**

– Control Theory

– Machine Learning: Meta‑learning concepts, PyTorch.

– Software: Python, chosen quadcopter simulation framework (e.g., PyBullet).

——————————————————————————————————–

**Key Reference**

– Finn, C., Abbeel, P., & Levine, S. (2017). Model‑Agnostic Meta‑Learning for Fast Adaptation of Deep Networks. In Proceedings of ICML (pp. 1126-1135).

Comment

If you are interested, please send your transcript (BSc and MSc) to the following email:

[email protected]

[email protected]

Professor(s)

Giancarlo Ferrari Trecate

Administration

Barbara Marie-Louise Frédérique Schenkel

The proposed project aims to explore and evaluate the use of meta-learning techniques in control systems.

Meta-learning focuses on enabling models to adapt quickly to new tasks with limited data by leveraging prior experience across related tasks. In this context, the project will focus on implementing algorithms that allow Neural Network-based controllers to efficiently generalize across a variety of control scenarios, adapting to new environments with minimal additional training.

The student will investigate the performance of established meta-learning methods, such as Model-Agnostic Meta-Learning (MAML) [1], in control tasks. By integrating these techniques with control-specific neural architectures [2] and evaluating them in simulation, the project will explore the potential of meta-learning for enhancing adaptability, robustness, and sample efficiency in dynamical systems. Simulations will be conducted on the Franka Robot, linking theoretical methods with practical, real-world application.

Main tasks:

-Review and understand relevant meta-learning algorithms

-Study Neural Network-based approaches to control

-Implement and evaluate selected approaches for control

-Perform simulations with the Franka Robot to assess real-world applicability

Required knowledge and tools:

-Control theory

-Machine learning

-Python and relevant libraries (PyTorch) for neural network implementation.

Reference papers:

– [1] C. Finn, P. Abbeel, and S. Levine, “Model-agnostic meta-learning for fast adaptation of deep networks,” in Proc. 34th Int. Conf. Machine Learning, Sydney, Australia, 2017, pp. 1126-1135.

– [2] L. Furieri, C. L. Galimberti, and G. Ferrari-Trecate, “Learning to boost the performance of stable nonlinear systems,” IEEE Open Journal of Control Systems, p. 342-357, 2024.

Comment

Contact:

[email protected]

[email protected]

Professor(s)

Giancarlo Ferrari Trecate, Christophe Salzmann

Administration

Barbara Marie-Louise Frédérique Schenkel

Previous work in the DECODE Lab (Furieri, 2024) proposed a method for learning stability-preserving nonlinear optimal controllers with neural networks. This framework, referred to as performance boosting (PB), has proven to be powerful for nonlinear and distributed control.

Recently, PB has been extended to switching controllers (Saccani, 2025) to maintain good performance in time-varying environments. While sufficient conditions for stability preservation during switches have been identified, the design of adaptive controllers for good practical performance remains an open challenge.

The goal of this project is to draw inspiration from online convex optimization (Hazan, 2016) to design new, theoretically sound, switching schemes for online adaptation. These schemes will have to be implemented and rigorously tested in a simulated setting for the control of a wind turbine.

Objectives:

[OB] Design new online switching schemes for performance-boosting controllers;

[SO1] Prove regret guarantees for the proposed switching schemes.

[SO2] Implement the proposed solutions in a simulated setting.

Required knowledge and tools:

– Control theory;

– Convex optimization;

– (Strong) experience with both Python and PyTorch.

Comment

Interested students should contact Sabri El Amrani ([email protected]) and Julien Pallage ([email protected]).

Professor(s)

Giancarlo Ferrari Trecate, Sabri El Amrani

Administration

Barbara Marie-Louise Frédérique Schenkel

External

DECODE

The objective of this project is to set up an Active Suspension system test bench and apply data-driven adaptive control techniques on it. Data acquisition and controller implementation would be done using LabVIEW and NI DAQs. The designed controller should be robust to the parametric uncertainties in the system and will be verified and validated on the test bench.

Professor(s)

Alireza Karimi, Vishnu Varadan

Administration

Barbara Marie-Louise Frédérique Schenkel

Site

https://www.epfl.ch/labs/la/, https://www.epfl.ch/labs/ddmac/