At the recent IEEE International Conference on Robotics and Automation (ICRA) in Brisbane, Australia, the Best Student Paper award went to ETH Zurich Autonomous Systems Laboratory (ASL)’s Miguel de la Iglesia Valls et al. for Design of an Autonomous Racecar: Perception, State Estimation and System Integration.
The ASL team’s racecar, dubbed “Flüela” after a Swiss mountain pass, won last year’s Formula Student Driverless competition in Hockenheimring, Germany. (ArXiv | GitHub) The car uses only onboard sensors and features a well-rounded system design that includes an EKF-based state estimation, LiDAR, and particle filter-based SLAM technologies. At Hockenheimring, “Flüela” completed 10 laps of a previously-unseen racetrack with performance comparable to human racecar drivers, and won the competition.
The racecar project aptly represents the ASL lab, which was missioned to create intelligent robots and systems that operate autonomously in complex and dynamic environments with a focus on ground, air, and water-based robotics applications, and complementary algorithms for perception, localization, abstraction, mapping, and path planning.
One of the lab’s recent advancements is in Simultaneous Localization and Mapping (SLAM), a technology that helps robots navigate their environments through either laser, sonar, or visual data input. Using a method called Open Keyframe-based Visual-Inertial SLAM (OKVIS), the robot can sync bits of visual information with sudden movements across small time-intervals. OKVIS was first published in The International Journal of Robotics Research in November 2014.
Founded at the École Polytechnique Fédérale de Lausanne in 1996, the ASL relocated to ETH in 2006. The lab has birthed numerous quirky, rolling, swimming, walking and flying robots over the years. It currently houses around 30 students. The lab has collaborated on projects with Google, Disney, or Microsoft, and just might spin off the next Swiss robotics unicorn.
The ICRA Best Student paper winners (and others nominated for Best Paper) were students of ASL Professor Roland Siegwart. Synced had the opportunity to speak with Professor Siegwarts in 2017, when he welcomed us at the lab at Leonhardstrasse 21 in Zurich. It was a fair May day, and although most students were attending conferences or had departed for overseas exchanges, many were planning on staying on at the lab and working through the summer.
Inside the lab, PhD student Raghav Khanna was busy fine-tuning a drone prototype that flies over crops fields to collect biomass development information, a project running in conjunction with cross-sector appliance Bosch. While in a hallway, another student took a screwdriver to the ASL quadrupedal robot ANYmal, garage-lab style.
ETH Zurich (Eidgenössische Technische Hochschule Zürich, or Swiss Federal Institute of Technology in Zurich) is Europe’s top STEM institution and one of the best universities in the world for robotics research, with eight labs and about 120 PhD students. The flying robots of Raffaello D’andrea of the Dynamics Systems and Control Lab have garnered a sensational three million YouTube views, while the work of Bradley Nelson on microrobotics and soft robots at the Institute of Robotics and Intelligent Systems has earned a broad international reputation.
The lab’s work lies at the nexus of hardware and software. A Chinese undergrad student tells Synced, “ETH is a very resourceful robotics research school. Hardware is expensive, but the school isn’t hesitant to put money into it. And many PhDs here have backgrounds in mechanical or electrical engineering.”
Professor Siegwart teaches autonomous mobile robotics at ETH Zurich and is the founding co-director of joint research and development center Wyss Zurich. His 36,286 Google citations rank him sixth among computer scientists in Switzerland.
A roboticist with a sober and welcoming Swiss temperament, Professor Siegwart explains that he has a “very functional view of machines,” as opposed to researchers who fascinate over androids, stressing that “robotics is a very very hard problem involving design and precision mechanics, perception, physical interaction, and intelligence.” He has repeatedly warned against the “hype and overpromising” he sees in AI and robotics.
“The physical world has many complications,” says Professor Siegwart. “If your smartphone crashes, you will buy another. But if an autonomous car breaks down, people may be killed. So far we are doing research in confined environments, but as soon as drones are flying over people, they should have failing modes and be extremely aware of people.”
ETH Zurich collaborates with many institutions, including Germany’s Max Planck Institute and Stanford University in the US. It also conducts joint research programs with the University of Zurich, with which it shares a campus.
Journalist: Meghan Han | Editor: Michael Sarazen
0 comments on “Robotic Innovations from ETH Zurich Autonomous Systems Laboratory”