The Soft Robotics Lab within the Institute of Robotics and Intelligent Systems at ETH Zurich is inviting applications for an open PhD position. We are looking for an excellent researcher to join and steer our research efforts in the control and machine learning of soft robots. Our lab's goal is to build, model, and control robots in a fundamentally different way, so that they become more flexible, dexterous, capable, and adapt better to their environment. The lab showcases its research in design and algorithms through real-world applications such as a robotic fish for underwater locomotion, soft robotic hands for pick-and-place tasks, and multi-segment arms for dynamic object manipulation.
Robots made of soft materials, so-called soft robots, well surpass the limited degrees of freedom that rigid systems possess and therefore potentially offer adaptable and inherently safe ways of achieving versatile forms in locomotion and manipulation. Soft robots can homogeneously combine actuation, sensing, and structure in one composite.
In our lab, we develop continuously deformable robots using compliant materials and intelligent sensors whose properties resemble those of living organisms. These soft robots can be used more flexibly and are safer than conventional ones when interacting with humans and surroundings. The vision of this project is to evolve soft robotic systems from the realm of proof-of-concept, one-off prototypes into capable and reliable machines that can morph to tackle challenging tasks in manipulation and locomotion.
One of the core challenges to overcome and realize the project's vision is the development of soft robotic "brains". Robotic “brains” are needed to fully exploit the unique properties of soft robots and make them useful for real-world applications. The controller within a robotic brain relies on a fast and representative model of the robotic “body” that is to be controlled. However, accurately modeling a deformable material under controlled actuation and/or interacting with the environment is complex and computationally slow because of the infinite dimensions of the material. Control approaches used for traditional robotic systems made of rigid links do not extend to the control of soft robots. Therefore, it is important to develop models and controllers for soft robots to enable them to perform challenging locomotion or manipulation tasks that surpass the capabilities of rigid robots.
You conceive and validate model-based controllers and reinforcement learning techniques that can perform difficult tasks requiring interaction with a robot’s surroundings. You advance the state of the art of the current modeling approaches used in the dynamic closed-loop control of soft continuum robots. With your approach, you encapsulate non-linear deformations of hyperelastic materials in real-time. Your model-based controllers become basic building blocks in a reinforcement learning framework. Your self-developed frameworks allow soft robots to learn advanced shapeshifting and autonomous manipulation capabilities for dexterous tasks. You continuously design, simulate, and test your control and learning algorithms on your deformable robots so that they dexterously interact with an inherently deformable world. You develop algorithms that make soft robots move in surprisingly dexterous ways.
Your system modeling and control approach is design-agnostic and user-friendly to facilitate broad adoption in the robotics community. You create an open-source control and learning framework for soft robots. The development of such a framework is facilitated through active collaborations with other researchers working on numerical simulations of physical systems such as SOFA and on toolboxes for dynamical systems modeling such as Drake.
Next to your research efforts, you are a passionate mentor and a patient educator. You are outstanding in mentoring students performing their semester projects or thesis works in our lab. You will also be involved in the teaching of graduate classes in soft robotics, modeling, controls, and machine learning.
You are extremely curious, highly driven, greatly independent, and you want to make a real difference with your research. You work best when you are in a team, you are respectful and thrive when you can collaborate with others and provide support. Through your prior experiences, you have shown your understanding in:
A foundational understanding of control and learning methods is essential for the successful completion of this proposed doctoral research program. It would be greatly advantageous if you are already proficient in C++ and Python. You collaborate well with your lab mates and other research labs, you are a resourceful asset for other researchers and students, and you can make efficient use of the provided support from R&D experts, scientists, engineers, and doctoral students with background in materials science, mechanical engineering, electrical engineering, computer science, and physics.
We look forward to receiving your online application with the following documents:
Please note that we exclusively accept applications submitted through our online application portal. Applications via email or postal services will not be considered. Applications missing required documents will not be considered. We will be in contact with you after 2-3 weeks following your application.
Further information about the Soft Robotics Lab can be found on our website www.srl.ethz.ch. Questions regarding the position should be directed by email to Prof. Katzschmann (no applications please).
|Titel||PhD Position in Control and Machine Learning of Soft Robots|
|Job location||Rämistrasse 101, 8006 Zurich|
|Veröffentlicht||August 3, 2020|
|Jobart||PhD/ Doktorand/in  |
|Fachbereiche||Algorithmen,   Künstliche Intelligenz,   Steuerungstechnik,   Robotertechnik,   Maschinelles Lernen,   Systems Engineering  |