Open Lab 28.11.2025
The Open Lab Day is an opportunity for students to learn about the ongoing projects. This year the Open Lab Day will be coorganised by IDSC and RobotX. There will be lectures as well as live demos.
REGISTRATION = WAITING LIST
You need to be registered to be able to enter the different venues. No registration - no entrance! Spaces are limited. We still have room in the lectures rooms for the talks.
Registration is officially closed as we’ve reached maximum capacity. If you would like to be put on the waiting list, please fill in the (waiting list) form below. If space becomes available, we will reach out to you and confirm your registration.
(*) The demos of IDSC will only last until 7:00 p.m. From 7:00 p.m. onwards, the IDSC staff will be present at the apero for students and researchers
You can find an overview of the talks with descriptions here
Visit the IDSC talks from 17:00-19:00
Strategic Cooperation in Multi-Agent Systems with Applications in Transportation Networks: This project explores strategic cooperation in multi-agent systems. Since self-interested behavior can lead to suboptimal outcomes, we design mechanisms that guide agents toward decisions that benefit both themselves and the overall system. The approach is validated in real-world transportation networks, demonstrating its potential to enhance the efficiency and sustainability of future large-scale systems.
Presenter: Mingjia He Co-design of Embodied Intelligence: Designing embodied intelligence systems is challenging due to the need to integrate diverse and interconnected hardware and software components, explore a vast design space, and manage trade-offs, as additional functionalities for the robot require more resources. The goal is to automate the design process by finding the optimal design choices for a robot, including perception, decision-making, body, sensors, and computing unit, tailored to a specific task while minimizing required resources such as cost, weight, power consumption, and computing need. To achieve this, we model the various autonomy components and their interconnections within a monotone co-design framework. Our project offers a range of research directions, including optimization, motion planning, perception, benchmarking, localization, and general robotics. This includes both theoretical work, such as developing new algorithms, and practical work, such as conducting experiments and case studies.
Presenters: Dejan Milojevic, Lorenzo Fiaschi Bridging Design, Compliance and Evaluation for Responsible Automation: As autonomous systems take on increasingly valuable and consequential tasks in daily life, it becomes essential to describe how these systems should behave and to ensure that they behave accordingly. However, defining behaviors remains an arduous, largely manual process, often prone to ambiguity, exploitability, and misrepresentation. At the same time, systems frequently lack the expressive capacity to encode behavioral specifications faithfully, leading to discrepancies between design intentions and real-world outcomes. Consequently, our research seeks to formalise and automate specification creation and their verification and evaluation. In this talk, we outline three interconnected directions towards responsible autonomy: Rule Design, developing practical formal frameworks mediating safety, efficiency, and compliance; Safety Certificates, providing verifiable guarantees of adherence within intended domains; and Robustness Evaluation, ensuring reliability beyond nominal conditions.
Presenters: Matteo Penlington, Yueshan Li Karma Economies: Towards Fair and Trustworthy Resource Allocation: What if fairness could be encoded, like an algorithm? Inspired by the idea of karma, our work develops a formal “give-and-take” framework for resource allocation systems.
Presenters: Kai Zhang, Antonia Holsten perfusionX - Redefining what’s possible in medicine: Every day, patients die waiting for a life-saving transplant. The shortage of viable donor organs is one of medicine’s most pressing challenges—and technology can change that. At perfusionX, we build a high-tech experimental ex vivo perfusion machine that keeps organs alive and functioning outside the human body. We’re developing the tools and treatments of tomorrow combining fluid dynamics, control engineering, thermodynamics, and biomedical science to simulate the body’s natural environment. If you want to apply your skills to a real-world, high-impact medical technology and help shape the future of transplantation, join us at the cutting edge of perfusion research.
Presenters: David Machacek, Ilyas Seçkin
*This talk will also include a separate demo in the RobotX hall. From Code to WebApp in 5 Minutes: In this talk, we show how the ETH SpinOff nihito.io can support you in going from a Python-based proof of concept (POC) to a minimum valuable product (MVP) in less than 5 minutes. We show practical examples of why and when this can be useful for student or research projects.
Presenter: Fabio Widmer CyberRunner: CyberRunner is an AI robot whose task is to learn how to play the popular and widely accessible labyrinth marble game. The labyrinth is a game of physical skill whose goal is to steer a marble from a given start point to the end point. In doing so, the player must prevent the ball from falling into any of the holes that are present on the labyrinth board.
Presenter: Thomas Bi
*This talk will also include a separate demo in the RobotX hall. Teaching a robot to master a board game using Reinforcement Learning: In this talk, we'll break down our journey —from simulating the game environment with physics engines to training RL agents. Discover how the robot learns optimal policies for skill acquisition through trial-and-error rewards. We'll share real-world implementation details, including hardware integration with robotic arms, computer vision for state detection, and strategies for overcoming challenges such as actuator mismatches and unpredictable physics.
Presenter: Aswin Ramachandran
*This talk will also include a separate demo in the RobotX hall. From Urban Mobility to Global Freight: Multi-Agent Control in Maritime Transportation Systems: Maritime transportation forms the backbone of the global economy, responsible for over 80% over global trade by volume and 50% of global trade value. In this talk we introduce key stakeholders, their interactions and explore "what if" scenarios to assess the impact of vessel fleet coordination schemes in tanker markets.
Presenter: Marc Albert Formula 1 Project: Efficient control algorithms and state-of-the-art off-line optimization tools are designed for the hybrid electric propulsion system of the Formula 1 car, to achieve the fastest possible lap-time while fulfilling the constraints on fuel and battery consumption.
Presenters: Joschua Wüthrich, Giona Fieni, Marc-Philippe Neumann
*This talk will also include a separate demo in the RobotX hall. Control, Learning and Optimization for Complex Systems:
Recent advancements in machine learning offer new opportunities for the design of high-performance controllers for many safety-critical applications, for example in power systems, robotics, biomedical applications, or manufacturing. At the same time, the use of learning in closed-loop control systems raises a number of challenges related to their safety and reliability, but also real-time computation requirements and data-efficiency. We develop the theory, methods and tools for advanced control, state estimation, and model learning with guaranteed satisfaction of critical safety constraints.
Contributors: Alexandre Didier, Marcell Bartos Biomedical Applications at ICS: The human body is a complex system that contains many different control loops to maintain life. When one of these feedback loops is not functioning properly, medical devices are used to replace their functionality accordingly. In our research projects, we aim to reproduce the dynamic behavior of healthy and diseased human subjects. Using in silico, in vitro, in vivo, and clinical simulations, we employ advanced computational methods to develop a comprehensive understanding of the human body, develop physiological controllers and implants, and provide targeted therapies.
Presenters: Marianne Schmid Daners, Martina Roncoroni, Marco Heim, Dominik Schulte
*This talk will also include a separate demo in the RobotX hall.
Presenter: Mingjia He Co-design of Embodied Intelligence: Designing embodied intelligence systems is challenging due to the need to integrate diverse and interconnected hardware and software components, explore a vast design space, and manage trade-offs, as additional functionalities for the robot require more resources. The goal is to automate the design process by finding the optimal design choices for a robot, including perception, decision-making, body, sensors, and computing unit, tailored to a specific task while minimizing required resources such as cost, weight, power consumption, and computing need. To achieve this, we model the various autonomy components and their interconnections within a monotone co-design framework. Our project offers a range of research directions, including optimization, motion planning, perception, benchmarking, localization, and general robotics. This includes both theoretical work, such as developing new algorithms, and practical work, such as conducting experiments and case studies.
Presenters: Dejan Milojevic, Lorenzo Fiaschi Bridging Design, Compliance and Evaluation for Responsible Automation: As autonomous systems take on increasingly valuable and consequential tasks in daily life, it becomes essential to describe how these systems should behave and to ensure that they behave accordingly. However, defining behaviors remains an arduous, largely manual process, often prone to ambiguity, exploitability, and misrepresentation. At the same time, systems frequently lack the expressive capacity to encode behavioral specifications faithfully, leading to discrepancies between design intentions and real-world outcomes. Consequently, our research seeks to formalise and automate specification creation and their verification and evaluation. In this talk, we outline three interconnected directions towards responsible autonomy: Rule Design, developing practical formal frameworks mediating safety, efficiency, and compliance; Safety Certificates, providing verifiable guarantees of adherence within intended domains; and Robustness Evaluation, ensuring reliability beyond nominal conditions.
Presenters: Matteo Penlington, Yueshan Li Karma Economies: Towards Fair and Trustworthy Resource Allocation: What if fairness could be encoded, like an algorithm? Inspired by the idea of karma, our work develops a formal “give-and-take” framework for resource allocation systems.
Presenters: Kai Zhang, Antonia Holsten perfusionX - Redefining what’s possible in medicine: Every day, patients die waiting for a life-saving transplant. The shortage of viable donor organs is one of medicine’s most pressing challenges—and technology can change that. At perfusionX, we build a high-tech experimental ex vivo perfusion machine that keeps organs alive and functioning outside the human body. We’re developing the tools and treatments of tomorrow combining fluid dynamics, control engineering, thermodynamics, and biomedical science to simulate the body’s natural environment. If you want to apply your skills to a real-world, high-impact medical technology and help shape the future of transplantation, join us at the cutting edge of perfusion research.
Presenters: David Machacek, Ilyas Seçkin
*This talk will also include a separate demo in the RobotX hall. From Code to WebApp in 5 Minutes: In this talk, we show how the ETH SpinOff nihito.io can support you in going from a Python-based proof of concept (POC) to a minimum valuable product (MVP) in less than 5 minutes. We show practical examples of why and when this can be useful for student or research projects.
Presenter: Fabio Widmer CyberRunner: CyberRunner is an AI robot whose task is to learn how to play the popular and widely accessible labyrinth marble game. The labyrinth is a game of physical skill whose goal is to steer a marble from a given start point to the end point. In doing so, the player must prevent the ball from falling into any of the holes that are present on the labyrinth board.
Presenter: Thomas Bi
*This talk will also include a separate demo in the RobotX hall. Teaching a robot to master a board game using Reinforcement Learning: In this talk, we'll break down our journey —from simulating the game environment with physics engines to training RL agents. Discover how the robot learns optimal policies for skill acquisition through trial-and-error rewards. We'll share real-world implementation details, including hardware integration with robotic arms, computer vision for state detection, and strategies for overcoming challenges such as actuator mismatches and unpredictable physics.
Presenter: Aswin Ramachandran
*This talk will also include a separate demo in the RobotX hall. From Urban Mobility to Global Freight: Multi-Agent Control in Maritime Transportation Systems: Maritime transportation forms the backbone of the global economy, responsible for over 80% over global trade by volume and 50% of global trade value. In this talk we introduce key stakeholders, their interactions and explore "what if" scenarios to assess the impact of vessel fleet coordination schemes in tanker markets.
Presenter: Marc Albert Formula 1 Project: Efficient control algorithms and state-of-the-art off-line optimization tools are designed for the hybrid electric propulsion system of the Formula 1 car, to achieve the fastest possible lap-time while fulfilling the constraints on fuel and battery consumption.
Presenters: Joschua Wüthrich, Giona Fieni, Marc-Philippe Neumann
*This talk will also include a separate demo in the RobotX hall. Control, Learning and Optimization for Complex Systems:
Recent advancements in machine learning offer new opportunities for the design of high-performance controllers for many safety-critical applications, for example in power systems, robotics, biomedical applications, or manufacturing. At the same time, the use of learning in closed-loop control systems raises a number of challenges related to their safety and reliability, but also real-time computation requirements and data-efficiency. We develop the theory, methods and tools for advanced control, state estimation, and model learning with guaranteed satisfaction of critical safety constraints.
Contributors: Alexandre Didier, Marcell Bartos Biomedical Applications at ICS: The human body is a complex system that contains many different control loops to maintain life. When one of these feedback loops is not functioning properly, medical devices are used to replace their functionality accordingly. In our research projects, we aim to reproduce the dynamic behavior of healthy and diseased human subjects. Using in silico, in vitro, in vivo, and clinical simulations, we employ advanced computational methods to develop a comprehensive understanding of the human body, develop physiological controllers and implants, and provide targeted therapies.
Presenters: Marianne Schmid Daners, Martina Roncoroni, Marco Heim, Dominik Schulte
*This talk will also include a separate demo in the RobotX hall.
Visit the demos of the different labs (IDSC only until 19.00)
RSL Transformer Robot: Transformer Robot can do it all: quadrupedal locomotion, bipedal locomotion, and dual-arm manipulation.
Contributors: Victor Klemm, Efe Ongan, Per Frivik, Elena Krasnova
https://www.wheeledleggedrobotics.com RSL Humanoids: Exhibition of walking controllers for various humanoid robotic platforms (Fourier N1/Unitree G1/Engine AI)
Contributors: Ioannis Dadiotis, Matthias Heyrman, Clemens Schwarke
“Thinking Inside the Scene” explores how robots can develop common sense by perceiving, interacting, and learning from within dynamic 3D environments — grounding understanding through spatial reasoning, affordances, and egocentric experience.
Contributors: Zuria Bauer, Tim Engelbracht, Boyang Sun, Alexey Gavryushin, Petar Lukovic, Panagiotis Minos, Rene Zurbrügg, Marwan Mohamed, Marc Pollefeys
https://cvg.ethz.ch/ ORCA Hand Dexterous Manipulation: The demo showcases the ability of dexterous manipulation of the ORCA hand, which is developed and open-sourced by SRL.
Contributors: Davide Liconti, Chenyu Yang
https://orca.ethz.ch/ Heavy Construction Robotics: The heavy construction robotics group at RSL develops autonomous and intelligent systems for heavy construction tasks. Our research combines advanced perception, control, and reinforcement learning to enable mobile robots and heavy machinery to operate safely and efficiently on complex construction sites across a wide variety of tasks such as autonomous earthworks and material handling, locomotion, and manipulation.
Contributors: Pol Eyschen, Claudio Canales, Marina Moreira, Fang Nan, Lorenzo Terenzi, Olga Vysotsaka, Lennart Werner ASL Aerial Robotics: We're building a novel generation of aerial robots - physically capable of doing work and capable of sensing and understanding their environments. At the booth, we'll show the current state of research and present our platform and team.
Contributors: Michael Pantic, Thomas Stastny, Mike Allenspach, Eugenio Cuniato, Christian Lanegger, Patrick Pfreundschuh, Lucy Harris, Mariano Biasio CyberRunner: CyberRunner is an AI robot whose task is to learn how to play the popular and widely accessible labyrinth marble game. The labyrinth is a game of physical skill whose goal is to steer a marble from a given start point to the end point. In doing so, the player must prevent the ball from falling into any of the holes that are present on the labyrinth board.
Contributor: Thomas Bi
https://www.cyberrunner.ai Perception & Navigation: At our booth we will showcase and discuss the algorithms and hardware developed at RSL for perception and navigation.
Contributors: William Talbot, Turcan Tuna, Pascal Roth, Julia Richter, Fan Yang
https://rsl.ethz.ch/research/researchtopics/perception-navigation.html Hydrocephalus test bench: Hydrocephalus is a condition in which excess cerebrospinal fluid accumulates in the brain, leading to neurological damage. Treatments such as ventriculoperitoneal (VP) shunts can alleviate symptoms, but optimizing their performance remains a challenge.The hybrid hydrocephalus test bench enables testing and validation of both traditional and state-of-the-art devices as well as of physiological controllers under realistic conditions, thereby reducing animal testing.The platform enables shunt testing under a variety of conditions, ranging from intracranial and abdominal pressure fluctuations to posture-induced pressure changes. It can replicate human dynamics but also allows us to mimic other individuals like sheep.Through the application of robotics, system modeling, and control engineering, this test bench offers an ethical, flexible, and reproducible alternative to animal testing and accelerates the development of next-generation hydrocephalus treatments.
Contributors: Karl Werner, Marianne Schmid Daners, Martina Roncoroni
https://ics.mavt.ethz.ch/research-projects/biomedical-applications.html
Formula 1 Project: Efficient control algorithms and state-of-the-art off-line optimization tools are designed for the hybrid electric propulsion system of the Formula 1 car, to achieve the fastest possible lap-time while fulfilling the constraints on fuel and battery consumption.
Contributors: Joschua Wüthrich, Giona Fieni, Marc-Philippe Neumann
https://idsc.ethz.ch/research-guzzella-onder/research-projects/Formula1.html MPC for autonomous racing and space applications: To simplify the time-consuming task of deploying new control and estimation algorithms, our laboratory presents Chronos and CRS. Chronos is a miniature radio-controlled racing car. CRS is a software framework for single and multi-agent control and robotics problems. The current CRS implementation includes classical control algorithms for autonomous racing, multi-agent autonomous racing control algorithms, collaborative control algorithms, and safety frameworks. With the upcoming projects, we are tackling several novel research directions: learning-based predictive control to unlock high-performance control, algorithms at the intersection between predictive control and reinforcement learning, vision-based control, and novel GNC algorithms for rocket control.
Contributors: Andrea Carron, Lukas Vogel, Sabrina Bodmer
https://ics.mavt.ethz.ch/research-projects/demonstrators.html Teaching a robot to master a board game using Reinforcement Learning: From simulating the game environment with physics engines to training RL agents —Discover how the robot learns optimal policies for skill acquisition through trial-and-error rewards. We'll share real-world implementation details, including hardware integration with robotic arms, computer vision for state detection, and strategies for overcoming challenges such as actuator mismatches and unpredictable physics.
Contributor: Aswin Ramachandran perfusionX - Redefining what’s possible in medicine: Every day, patients die waiting for a life-saving transplant. The shortage of viable donor organs is one of medicine’s most pressing challenges—and technology can change that. At perfusionX, we build a high-tech experimental ex vivo perfusion machine that keeps organs alive and functioning outside the human body. We’re developing the tools and treatments of tomorrow combining fluid dynamics, control engineering, thermodynamics, and biomedical science to simulate the body’s natural environment. If you want to apply your skills to a real-world, high-impact medical technology and help shape the future of transplantation, join us at the cutting edge of perfusion research.
Contributors: David Machacek, Ilyas Seckin
https://www.perfusionx.ch/ Smartphone-Assisted River Analysis - SARA: Project SARA is the next iteration of our bridge-mounted camera system to detect waste in rivers and measure pollution. While some data exists, it is sparse, inconclusive, and often only regional. Therefore, we want to build a monitoring system for large-scale unified monitoring. The system aims to be deployable worldwide and provide valuable river pollution data for researchers, NGOs, and government institutions. Our novelty is that we will use a smartphone as the core component of the monitoring system. Today's smartphones are compact, relatively inexpensive, and equipped with a good camera and high computing power. The project's main tasks are the development of an app for river waste detection and the system hardware for scalable and long-term environmental monitoring at multiple bridge locations. The app also allows field workers in developing countries to use their phones for waste quantification.
Contributors: Emre Elbir, Mateo Haro, Leander Joseph
https://sara-system.ethz.ch/ Mobile Robotics Lab: We will have a demo of perception on humanoid robots and semantic mapping on drones, showcasing the Mobile Robotics Lab's research on visual SLAM and scene understanding.
Contributors: Micha Bosshart, Helen Oleynikova, Alexander Hansson, Joshua Näf, Leonard Freissmuth, Shi Chen
https://mrl.ethz.ch/ RSL Mobile manipulation: The ALMA (Articulated Locomotion and MAnipulation) Legged Manipulator Robot: explore the capability of mobile manipulation with a quadrupedal robot mounted with an arm.
Contributors: Mattia Risiglione, Tianxu An
https://rsl.ethz.ch/research/researchtopics/mobile-manipulation.html magnecko is a magnetic quadrupedal climbing robot which aims to navigate complex three-dimensional structures to perform inspection tasks. We will show magneto in different development stages with the latest version of magnecko.
Contributors: Nicolas Faesch, Nicolas Hürlimann, Connor Flynn, Michael Haslinger, Stefan Leuthard, Timo Eugster
https://magnecko.ethz.ch/ Space Robotics and Hardware Development at RSL: LunarLeaper is a robotic mission concept designed to access and explore lunar lava tubes, using agile legged mobility to overcome hazardous terrain and enable in-depth investigation of these subsurface environments. Come learn about this project and other space and hardware developments at RSL!
Contributors: Joe Church, Elena Krasnova, Marco Trentini, Robert Baines
https://www.lunarleaper.space/ Flexion is building the autonomy stack for humanoid robots - from command to control, from manipulation to locomotion, across any hardware and task. Leveraging the power of simulation and reinforcement learning, their software scales to the real world with minimal human involvement.
https://flexion.ai/
RIVR is a Swiss robotics company pioneering Physical Al and robotic solutions to revolutionize last-mile delivery, giving 1 human the power of1000. Through the combination of artificial neural networks and innovative robot designs with wheels and legs, RIVR aims to enhance efficiency, sustainability, and scalability in last-mile delivery. Founded as Swiss-Mile, the company rebranded to RIVR in 2025 to better reflect its mission of driving the future of intelligent robotics.
https://www.rivr.ai/ RSL Navigation (Real + Simulation)
Contributors: Fan Yang, Per Frivik We will present RAVI – Robot Assistant for Visual Independence – the hardware palette and software stack developed during the last 7 months for the Neura Robotics Challenge. The hardware parts consist of a small electronic wearable, a custom gripper for the robot arm and tactile sensors. We will also show the robot arm in simulation (Isaac Sim) to emulate the real physical system and demonstrate the software stack. The simulated as well as real robot arm are controlled via ROS2 and integrated into a network where an agentic ai can take control over the system, orchestrating tasks.
Contributors: Felix Schempp, Ugur Kafkasyali, Camila Schmidt
https://www.cafelabs.org/
Contributors: Victor Klemm, Efe Ongan, Per Frivik, Elena Krasnova
https://www.wheeledleggedrobotics.com RSL Humanoids: Exhibition of walking controllers for various humanoid robotic platforms (Fourier N1/Unitree G1/Engine AI)
Contributors: Ioannis Dadiotis, Matthias Heyrman, Clemens Schwarke
“Thinking Inside the Scene” explores how robots can develop common sense by perceiving, interacting, and learning from within dynamic 3D environments — grounding understanding through spatial reasoning, affordances, and egocentric experience.
Contributors: Zuria Bauer, Tim Engelbracht, Boyang Sun, Alexey Gavryushin, Petar Lukovic, Panagiotis Minos, Rene Zurbrügg, Marwan Mohamed, Marc Pollefeys
https://cvg.ethz.ch/ ORCA Hand Dexterous Manipulation: The demo showcases the ability of dexterous manipulation of the ORCA hand, which is developed and open-sourced by SRL.
Contributors: Davide Liconti, Chenyu Yang
https://orca.ethz.ch/ Heavy Construction Robotics: The heavy construction robotics group at RSL develops autonomous and intelligent systems for heavy construction tasks. Our research combines advanced perception, control, and reinforcement learning to enable mobile robots and heavy machinery to operate safely and efficiently on complex construction sites across a wide variety of tasks such as autonomous earthworks and material handling, locomotion, and manipulation.
Contributors: Pol Eyschen, Claudio Canales, Marina Moreira, Fang Nan, Lorenzo Terenzi, Olga Vysotsaka, Lennart Werner ASL Aerial Robotics: We're building a novel generation of aerial robots - physically capable of doing work and capable of sensing and understanding their environments. At the booth, we'll show the current state of research and present our platform and team.
Contributors: Michael Pantic, Thomas Stastny, Mike Allenspach, Eugenio Cuniato, Christian Lanegger, Patrick Pfreundschuh, Lucy Harris, Mariano Biasio CyberRunner: CyberRunner is an AI robot whose task is to learn how to play the popular and widely accessible labyrinth marble game. The labyrinth is a game of physical skill whose goal is to steer a marble from a given start point to the end point. In doing so, the player must prevent the ball from falling into any of the holes that are present on the labyrinth board.
Contributor: Thomas Bi
https://www.cyberrunner.ai Perception & Navigation: At our booth we will showcase and discuss the algorithms and hardware developed at RSL for perception and navigation.
Contributors: William Talbot, Turcan Tuna, Pascal Roth, Julia Richter, Fan Yang
https://rsl.ethz.ch/research/researchtopics/perception-navigation.html Hydrocephalus test bench: Hydrocephalus is a condition in which excess cerebrospinal fluid accumulates in the brain, leading to neurological damage. Treatments such as ventriculoperitoneal (VP) shunts can alleviate symptoms, but optimizing their performance remains a challenge.The hybrid hydrocephalus test bench enables testing and validation of both traditional and state-of-the-art devices as well as of physiological controllers under realistic conditions, thereby reducing animal testing.The platform enables shunt testing under a variety of conditions, ranging from intracranial and abdominal pressure fluctuations to posture-induced pressure changes. It can replicate human dynamics but also allows us to mimic other individuals like sheep.Through the application of robotics, system modeling, and control engineering, this test bench offers an ethical, flexible, and reproducible alternative to animal testing and accelerates the development of next-generation hydrocephalus treatments.
Contributors: Karl Werner, Marianne Schmid Daners, Martina Roncoroni
https://ics.mavt.ethz.ch/research-projects/biomedical-applications.html
Formula 1 Project: Efficient control algorithms and state-of-the-art off-line optimization tools are designed for the hybrid electric propulsion system of the Formula 1 car, to achieve the fastest possible lap-time while fulfilling the constraints on fuel and battery consumption.
Contributors: Joschua Wüthrich, Giona Fieni, Marc-Philippe Neumann
https://idsc.ethz.ch/research-guzzella-onder/research-projects/Formula1.html MPC for autonomous racing and space applications: To simplify the time-consuming task of deploying new control and estimation algorithms, our laboratory presents Chronos and CRS. Chronos is a miniature radio-controlled racing car. CRS is a software framework for single and multi-agent control and robotics problems. The current CRS implementation includes classical control algorithms for autonomous racing, multi-agent autonomous racing control algorithms, collaborative control algorithms, and safety frameworks. With the upcoming projects, we are tackling several novel research directions: learning-based predictive control to unlock high-performance control, algorithms at the intersection between predictive control and reinforcement learning, vision-based control, and novel GNC algorithms for rocket control.
Contributors: Andrea Carron, Lukas Vogel, Sabrina Bodmer
https://ics.mavt.ethz.ch/research-projects/demonstrators.html Teaching a robot to master a board game using Reinforcement Learning: From simulating the game environment with physics engines to training RL agents —Discover how the robot learns optimal policies for skill acquisition through trial-and-error rewards. We'll share real-world implementation details, including hardware integration with robotic arms, computer vision for state detection, and strategies for overcoming challenges such as actuator mismatches and unpredictable physics.
Contributor: Aswin Ramachandran perfusionX - Redefining what’s possible in medicine: Every day, patients die waiting for a life-saving transplant. The shortage of viable donor organs is one of medicine’s most pressing challenges—and technology can change that. At perfusionX, we build a high-tech experimental ex vivo perfusion machine that keeps organs alive and functioning outside the human body. We’re developing the tools and treatments of tomorrow combining fluid dynamics, control engineering, thermodynamics, and biomedical science to simulate the body’s natural environment. If you want to apply your skills to a real-world, high-impact medical technology and help shape the future of transplantation, join us at the cutting edge of perfusion research.
Contributors: David Machacek, Ilyas Seckin
https://www.perfusionx.ch/ Smartphone-Assisted River Analysis - SARA: Project SARA is the next iteration of our bridge-mounted camera system to detect waste in rivers and measure pollution. While some data exists, it is sparse, inconclusive, and often only regional. Therefore, we want to build a monitoring system for large-scale unified monitoring. The system aims to be deployable worldwide and provide valuable river pollution data for researchers, NGOs, and government institutions. Our novelty is that we will use a smartphone as the core component of the monitoring system. Today's smartphones are compact, relatively inexpensive, and equipped with a good camera and high computing power. The project's main tasks are the development of an app for river waste detection and the system hardware for scalable and long-term environmental monitoring at multiple bridge locations. The app also allows field workers in developing countries to use their phones for waste quantification.
Contributors: Emre Elbir, Mateo Haro, Leander Joseph
https://sara-system.ethz.ch/ Mobile Robotics Lab: We will have a demo of perception on humanoid robots and semantic mapping on drones, showcasing the Mobile Robotics Lab's research on visual SLAM and scene understanding.
Contributors: Micha Bosshart, Helen Oleynikova, Alexander Hansson, Joshua Näf, Leonard Freissmuth, Shi Chen
https://mrl.ethz.ch/ RSL Mobile manipulation: The ALMA (Articulated Locomotion and MAnipulation) Legged Manipulator Robot: explore the capability of mobile manipulation with a quadrupedal robot mounted with an arm.
Contributors: Mattia Risiglione, Tianxu An
https://rsl.ethz.ch/research/researchtopics/mobile-manipulation.html magnecko is a magnetic quadrupedal climbing robot which aims to navigate complex three-dimensional structures to perform inspection tasks. We will show magneto in different development stages with the latest version of magnecko.
Contributors: Nicolas Faesch, Nicolas Hürlimann, Connor Flynn, Michael Haslinger, Stefan Leuthard, Timo Eugster
https://magnecko.ethz.ch/ Space Robotics and Hardware Development at RSL: LunarLeaper is a robotic mission concept designed to access and explore lunar lava tubes, using agile legged mobility to overcome hazardous terrain and enable in-depth investigation of these subsurface environments. Come learn about this project and other space and hardware developments at RSL!
Contributors: Joe Church, Elena Krasnova, Marco Trentini, Robert Baines
https://www.lunarleaper.space/ Flexion is building the autonomy stack for humanoid robots - from command to control, from manipulation to locomotion, across any hardware and task. Leveraging the power of simulation and reinforcement learning, their software scales to the real world with minimal human involvement.
https://flexion.ai/
RIVR is a Swiss robotics company pioneering Physical Al and robotic solutions to revolutionize last-mile delivery, giving 1 human the power of1000. Through the combination of artificial neural networks and innovative robot designs with wheels and legs, RIVR aims to enhance efficiency, sustainability, and scalability in last-mile delivery. Founded as Swiss-Mile, the company rebranded to RIVR in 2025 to better reflect its mission of driving the future of intelligent robotics.
https://www.rivr.ai/ RSL Navigation (Real + Simulation)
Contributors: Fan Yang, Per Frivik We will present RAVI – Robot Assistant for Visual Independence – the hardware palette and software stack developed during the last 7 months for the Neura Robotics Challenge. The hardware parts consist of a small electronic wearable, a custom gripper for the robot arm and tactile sensors. We will also show the robot arm in simulation (Isaac Sim) to emulate the real physical system and demonstrate the software stack. The simulated as well as real robot arm are controlled via ROS2 and integrated into a network where an agentic ai can take control over the system, orchestrating tasks.
Contributors: Felix Schempp, Ugur Kafkasyali, Camila Schmidt
https://www.cafelabs.org/
