Student Projects

Currently, the following student projects are available. Please contact the responsible supervisor and apply with your CV and transcripts.

In case you have project ideas related to any of these projects, take the opportunity and propose your own project!

We also offer Master Thesis and or exchange semesters at reknown universities around the world. Please contact us in case of interest!

Studies on Mechatronics

We offer students also to conduct their Studies on Mechatronics at our lab. In general, we recommend to do the Studies on Mechatronics in combination with the Bachelor Thesis, either as prepartory work the semester beofre or as extended study in parallel. If you want to do it independently, yiou can find prroposed projets also in the list below. Please directly apply with corresponsponding supervisor.

ETH Zurich uses SiROP to publish and search scientific projects. For more information visit sirop.org.

Building with Rocks: Autonomous Construction with Heavy Machinery under Partial Observability

Autonomous heavy machinery has the potential to transform construction, enabling new sustainable building paradigms such as dry-stone walls, irregular masonry, and rubble-based structures that today require costly skilled labor. Unlike standardized bricks or prefabricated blocks, natural rocks are irregular, heavy, and unique: their geometry, mass distribution, and graspable surfaces are unknown a priori, and during construction they must be perceived, grasped, and stacked under significant partial observability, rocks may be partially buried, mutually occluded, or only visible from a few viewpoints reachable by the machine. Keywords: machine learning, deep learning, reinforcement learning, diffusion, construction robotics, space robotics

Keywords

machine learning, deep learning, reinforcement learning, diffusion, construction robotics, space robotics

Labels

Semester Project , Master Thesis

Description

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-06-02

Organization Robotic Systems Lab

Hosts Canales Claudio , Gouveia Marina

Topics Information, Computing and Communication Sciences , Engineering and Technology

Bidexterous Manipulation with Tactile Feedback

Humanoid robots are increasingly equipped with dexterous hands that enable them to interact with objects in more flexible, human-like ways. However, robust manipulation remains challenging, especially when relying only on external perception. The BrainCo Revo 2 hands provide tactile sensing on the fingertips, opening the possibility of combining touch with exteroceptive observations for object interaction. In this project, the student will develop a manipulation pipeline for tabletop tasks using the G1 humanoid and BrainCo dexterous hands. The project will study how combining exteroceptive perception with tactile feedback can improve manipulation performance. The final goal is to build and evaluate a first visuo-tactile manipulation pipeline for a bimanual dexterous platform, with a focus on understanding how tactile feedback can improve object-centric manipulation.

Keywords

Humanoid Manipulation, Dexterous Hands, Tactile Sensing, Visuo-Tactile Fusion, Robot Learning

Labels

Master Thesis

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-05-30 , Earliest start: 2026-06-01 , Latest end: 2027-03-01

Organization Robotic Systems Lab

Hosts Bhardwaj Arjun

Topics Information, Computing and Communication Sciences , Engineering and Technology

RL Finetuning for Generalized Locomotion

This project investigates the potential of reinforcement learning (RL) fine-tuning to develop a single, universal locomotion policy for quadruped or humanoid robots. Building on prior work in multi-terrain skill synthesis [1], we will probe the limits of generalization by systematically fine-tuning on an ever-expanding set of diverse environments. This incremental approach will test the hypothesis that a controller can learn to robustly navigate a vast range of terrains. As a potential extension, procedural terrain generation may be used to automatically create novel challenges, pushing the boundaries of policy robustness.

Keywords

Reinforcement Learning, Locomotion, Quadruped, Humanoid

Labels

Master Thesis

Description

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-05-29 , Earliest start: 2026-02-01

Applications limited to ETH Zurich

Organization Robotic Systems Lab

Hosts Schwarke Clemens , He Junzhe

Topics Information, Computing and Communication Sciences

High-speed locomotion learning on the Unitree A2 Quadruped

This project aims to develop robust high-speed locomotion for the Unitree A2 quadrupedal robot. The goal is to train and deploy omnidirectional locomotion policies capable of reaching up to 5 m/s while respecting thermal and actuator limits. The student will combine literature research, reinforcement learning in simulation, thermal dynamics modeling, and real-world deployment on the robot.

Keywords

Legged robotics, quadrupedal locomotion, reinforcement learning, PPO, sim-to-real transfer, high-speed locomotion, thermal modeling, actuator constraints, Unitree A2, robot learning

Labels

Semester Project , Master Thesis

Description

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-05-29 , Earliest start: 2026-06-28 , Latest end: 2027-03-01

Organization Robotic Systems Lab

Hosts Bjelonic Filip

Topics Information, Computing and Communication Sciences , Engineering and Technology

Propose Your Own Robotics Project

This project invites you to step into the role of an innovator, encouraging you to identify challenges you are passionate about within the field of robotics. Rather than working on predefined problems, you will have the freedom to propose your own project ideas, address real-world issues, or explore cutting-edge topics. This project allows you to define your own research journey.

Keywords

Robotics, Research

Labels

Semester Project , Bachelor Thesis , Master Thesis

Description

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-05-29 , Earliest start: 2025-01-27

Organization Robotic Systems Lab

Hosts Schwarke Clemens , Bjelonic Filip , Klemm Victor

Topics Information, Computing and Communication Sciences

Online Semantic Gaussian Splatting and RL for Mobile Robots

In this project, we aim to extend a Gaussian Splatting framework by integrating online semantics and build an RL framework that draws from both geometry and semantics in real time. We would like to explore how these semantically enriched representations can improve generalizable robot learning and sim-to-real transfer on agile mobile robots.

Keywords

Gaussian Splatting, Semantics, VFM, Reinforcement Learning, 3D Vision

Labels

Master Thesis

Description

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-05-27 , Earliest start: 2026-06-08

Organization Robotic Systems Lab

Hosts Gawel Abel

Topics Information, Computing and Communication Sciences

Emergent Acoustic Communication in Robot Groups

Can robots develop their own communication system through sound? We are looking for motivated students to join an interdisciplinary research project at the intersection of robotics, AI, acoustic signaling, and collective behavior.

Keywords

Emergent acoustic communication, Robot Learning, multi-agent communication.

Labels

Semester Project , Master Thesis

Description

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-05-16 , Earliest start: 2026-05-15

Organization Robotic Systems Lab

Hosts An Tianxu , Wang Shengzhi

Topics Information, Computing and Communication Sciences , Engineering and Technology

Hybrid Triangle and Gaussian Splatting

Photorealistic simulation and high-speed rendering are the basis of many modern robotics projects, but we are currently trapped in a binary: the rigid, memory-efficient world of triangle meshes versus the high-fidelity, computationally expensive world of neural representations. We propose a Hybrid Geometry Engine that abandons this "one-size-fits-all" approach. By representing static, flat architectural surfaces as optimized triangle meshes and complex 3D objects or volumetrics as 3D Gaussians, this project seeks to unlock a new paradigm for large-scale digital twins and robotic simulation.

Keywords

Computer Graphics 3D Gaussian Splatting Rendering

Labels

Master Thesis

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project... 

Published since: 2026-05-14 , Earliest start: 2026-05-18

Applications limited to ETH Zurich

Organization Robotic Systems Lab

Hosts Wilder-Smith Max

Topics Information, Computing and Communication Sciences

3D Scene Completion in Gaussian Splatting

Robots rarely get the luxury of a perfect map. During exploration, they are constantly fighting against "unknowables" such as occlusions, shadows, and unvisited corners that leave gaps in their understanding of the world. This project aims to address this by implementing 3D Gaussian Scene Completion, where the system proactively fills unobserved regions with learned geometric priors.

Keywords

3D Gaussian Splatting Scene reconstruction Scene completion 3D Graphics

Labels

Master Thesis

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project... 

Published since: 2026-05-14 , Earliest start: 2026-05-18

Applications limited to ETH Zurich

Organization Robotic Systems Lab

Hosts Wilder-Smith Max

Topics Information, Computing and Communication Sciences

Fine-Tuning Multimodal Reasoning Models for Affordance and Manipulativity Prediction in Robotic Manipulation

This project investigates whether reasoning-oriented multimodal models such as Gemma 4 can be fine-tuned to predict task-relevant affordances and manipulability for robotic manipulation. Given visual observations and high-level instructions, the model will identify actionable object parts, infer their functional suitability, and estimate how effectively a robot can interact with them using its own embodiment. Predictions may include affordance classes, contact regions, or manipulability scores learned from demonstrations, simulation, or robot execution data. These outputs will support interaction selection directly or guide downstream planners and control policies. The approach will be evaluated in simulation and on real robotic platforms to improve task-aware manipulation success in unstructured environments.

Keywords

affordance prediction; robotic manipulation; manipulability; multimodal reasoning; vision-language models; task-aware interaction; object part affordances; embodied AI; contact region prediction; motion planning; robot learning; high-level instruction understanding

Labels

Master Thesis

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project... 

Published since: 2026-05-10 , Earliest start: 2026-05-06 , Latest end: 2027-04-08

Applications limited to ETH Zurich , University of Zurich

Organization Robotic Systems Lab

Hosts Zurbrügg René , Werner Lennart

Topics Engineering and Technology

Automated calibration of body-scale contact sensors

Goal: Create a repeatable, precise, and high-throughput robotic experimental setup for characterizing a custom contact sensor. This project is part of a broader effort to enable robots to better perceive whole-body contact and execute dynamic, contact-rich tasks.

Keywords

Robotics, Tactile Sensing, Machine Learning, Automated Calibration, RL, AI, Mechanical Engineering, Manipulation

Labels

Semester Project

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project... 

Published since: 2026-05-07 , Earliest start: 2026-05-11 , Latest end: 2026-12-31

Applications limited to ETH Zurich

Organization Robotic Systems Lab

Hosts Baines Robert , Cramariuc Andrei

Topics Information, Computing and Communication Sciences , Engineering and Technology

Scene graphs meet implicit representations

Explore novel, hierarchical spatial memory representations for flexible robot autonomy

Keywords

Hierarchical scene graphs, implicit representations, RL

Labels

Master Thesis

Description

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-03-19 , Earliest start: 2026-03-23

Organization Robotic Systems Lab

Hosts Kneip Laurent

Topics Information, Computing and Communication Sciences , Engineering and Technology

LLM-based scene graph construction and navigation

Use LLM priors for efficient scene graph construction and navigation in urban environments

Keywords

Open vocabulary scene graphs, urban navigation, planning

Labels

Master Thesis

Description

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-03-19 , Earliest start: 2026-03-23

Organization Robotic Systems Lab

Hosts Kneip Laurent

Topics Information, Computing and Communication Sciences , Engineering and Technology

Learning-based robot state estimation

Achieve high-quality state estimation on any platform via conditioned neural representations

Keywords

Learning-based state estimation

Labels

Master Thesis

Description

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-03-19 , Earliest start: 2026-03-23

Organization Robotic Systems Lab

Hosts Kneip Laurent

Topics Information, Computing and Communication Sciences , Engineering and Technology

Exploring Exotic Wheels For Wheeled-Legged Robots

Wheeled-legged robots combine the advantages of wheels, i.e., efficient high-speed locomotion, with obstacle negotiation and rough terrain capabilities of legs. In this study, different exotic wheel concepts should be researched, evaluated, and possibly implemented on a real robot to potentially boost efficiency, manipulation, and functionality.

Keywords

Wheeled-Legged Robots, Wheels, Robotics, Mechanical and Hardware Design, Manipulation, Tires

Labels

Semester Project , Bachelor Thesis , Master Thesis

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project... 

Published since: 2026-02-18

Organization Robotic Systems Lab

Hosts Klemm Victor

Topics Information, Computing and Communication Sciences , Engineering and Technology

Hierarchical Reasoning Models (HRM) for Robotics

How to properly deal with long-horizon tasks is still an unsolved problem in robot learning. Recent advances in reasoning models present Hierarchical Reasoning Models (HRM), inspired by the hierarchical and multi-timescale processing in the human brain. This fairly open project should implement and test state of the art solutions for long horizon planning that require reasoning in the context of robotics, especially robotics motion control.

Keywords

Robot learning, hierarchical reasoning, machine intelligence, long-horizon planning

Labels

Semester Project , Bachelor Thesis , Master Thesis

Description

Work Packages

Requirements

Contact Details

More information

Open this project... 

Published since: 2026-02-18

Organization Robotic Systems Lab

Hosts Klemm Victor

Topics Information, Computing and Communication Sciences , Engineering and Technology

Student Theses at RAI

The Robotic Systems Lab offers a number of student projects at RAI or in collaboration with RAI. Please check the individual projects and directly apply to the supervisors.

Note on plagiarism

We would like to suggest every student, irrespective of the type of project (Bachelor, Semester, Master, ...), to make himself/herself familiar with ETH rules regarding plagiarism

JavaScript has been disabled in your browser