Stelian Coros
ETH Zurich

I am an Assistant Professor in the Department of Computer Science at ETH Zurich, where I lead the Computational Robotics Lab (CRL). I am also an Adjunct Professor in the Robotics Institute at Carnegie Mellon University. I received my PhD in Computer Science from the University of British Columbia in 2011.

I envision a future where robots will make the world a better place by becoming skilled co-workers and trusted social companions. To help make this vision come true, the overarching goal of my research is to enable robots to understand the physical world, and therefore the physical implication of their actions. This is a crucial step in the quest for robot intelligence, and it is what will enable robots to go wherever we go, to use tools and manipulate physical systems as effectively as we do, to assist us with a variety of tasks at work and at home. More specifically, I focus on a set of fundamental problem domains that constitute the pillars of my research program: robotic mobility, dexterous manipulation, and the intricate interplay between form and function for compliant systems. For further information, please refer to my group's homepage, crl.ethz.ch, and see the topics below:

Email: scoros@inf.ethz.ch


Research Interests::


Future robot ecosystems: Whether it is to augment human capabilities in the workplace or to help with chores around our homes, future generations of robots will need the skills to perform an increasingly diverse array of tasks. Consequently, I seek to develop the computational underpinnings to 1) endow robots with the intelligence necessary to dexterously manipulate complex physical systems, 2) establish a systematic way of reasoning about motor skills that combine locomotion and manipulation capabilities, and 3) enable heterogeneous robot-robot and robot-human teams to collaboratively undertake complex tasks.
Next-gen robots: In the animal kingdom, form and function are inseparably intertwined, and nature abounds with fascinating examples that illustrate how anatomical structures and motion capabilities are designed in concert. Soft tissues, in particular, are an integral part of the design of every biomechanical system, defining the performance, efficiency, robustness and safety of its movements. One of my long-term goals is to establish a systematic way of exploring this mechanical side of intelligence in order to create new breeds of robots that exploit compliant materials as effectively as living creatures do.
Computational fabrication: 3D Printing is unmatched in its ability to create complex geometric structures, it employs an ever-expanding range of materials and it can create one-off parts at virtually no extra cost. These exciting new capabilities are paving the way to a shift from mass production to personalized design and fabrication. However, they also introduce significant research challenges: the vast space of design possibilities far exceeds our current ability to create content for digital fabrication. To overcome this technological barrier, I am developing novel CAD tools powered by physics-based design.
Motor control models: Humans and animals move with remarkable skill, grace and agility. And while we devote little thought to moving around, even mundane tasks like walking require a tremendously complex interplay of sensory information processing, motion planning, and coordinated muscle control. One of my main research goals is to study the mathematical, biomechanical and motor-learning principles required to reproduce the wide range of motions seen in nature.



Honors and Awards:


National Science Foundation: Research Initiation Initiative Award, 2016
Intel Early Career Faculty Award, 2015
Alain Fournier Award: best Canadian Computer Graphics Dissertation, 2010
NSERC Canada Graduate Scholarship, 2006

Professional Activities: Technical program chair


ACM Symposium on Computational Fabrication, 2017.
ACM Symposium on Computer Animation, 2015.

Press (selection)


IEEE Spectrum: Clever Modular Robots Turn Legs Into Arms on Demand
Eurekalert: Bendy bots use twisty telescopic limbs
Eurekalert: CMU's interactive tool helps novices and experts make custom robots
Makezine: Design Tool for 3D-Printable Robots from Disney Research
3ders: Disney Research makes 3D printable robotic design easier than ever with interactive design tool
IEEE Spectrum: Disney Software Makes It Easy to Design and Print Custom Walking Robots
Wired UK: Disney Research helps novices 3D print robots from scratch
New Scientist: 3D print extra bits for old objects to help extend their life
Discovery Channel: Daily Planet Interview: Designing 3D printable robots
Wired: Disney Infinity STAR WARS reinvents the classic AT-AT takedown
Gizmag: Disney Research software makes mechanizing characters easy
Gizmodo: Animatronics Could Go Mainstream Thanks to Disney's Latest Program
Wired: Disney Research: computational design of mechanical characters
Wired UK: Disney software simplifies creation of gear-driven automata
The Engineer UK: Mechanical motion added to 3D-printed creations
3D Printing Industry: Disney Develops Method to Simplify Animatronics with 3D Printable Parts
Phys.org: Software systems add motion to physical characters
New Scientist: Virtual walkers lead the way for robots