Robohub.org
 

Team CERBERUS wins the DARPA Subterranean Challenge


by
01 October 2021



share this:

The DARPA Subterranean Challenge planned to develop novel approaches to rapidly map, explore and search underground environments in time-sensitive operations critical for the civilian and military domains alike. In the Final Event, DARPA designed an environment involving branches representing all three challenges of the “Tunnel Circuit”, the “Urban Circuit” and the “Cave Circuit”. Robots had to explore, search for objects (“artifacts”) of interest, and report their accurate location within underground tunnels, infrastructure similar to a subway, and natural caves and paths with extremely confined geometries, tough terrain, and severe visual degradation (including dense smoke).

Team CERBERUS deployed a diverse set of robots with the prime systems being four ANYmal C legged systems. In the Prize Round of the Final Event, the team won the competition and scored 23 points by correctly detecting and localizing 23 of 40 of the artifacts DARPA had placed inside the environment. The second team, “CSIRO Data61” also scored 23 points but reported the last artifact with a slight further delay to DARPA thus the tiebraker was in favor of Team CERBERUS. The third team, “MARBLE” scored 18 points.

The DARPA Subterranean Challenge was one of the rare types of global robotic competition events pushing the frontiers for resilient autonomy and calling teams to develop novel and innovative solutions with the capacity to help critical sectors such as search and rescue personnel and the industry in domains such as mining and beyond. The level of achievement of Team CERBERUS is best understood by looking at all the competitors in the “Systems Competition” of the Final Event. The participating teams including members from top international institutions, namely:

  • CERBERUS (Score = 23): University of Nevada, Reno, ETH Zurich, NTNU, University of California Berkeley, Oxford Robotics Institute, Flyability, Sierra Nevada Corporation
  • CSIRO Data61 (Score = 23): CSIRO, Emesent, Georgia Institute of Technology
  • MARBLE (Score = 18): University of Colorado Boulder, University of Colorado Denver, Scientific Systems Company, University of California Santa Cruz
  • Explorer (Score = 17): Carnegie Mellon University, Oregon State University
  • CoSTAR (Score = 13): NASA Jet Propulsion Laboratory, California Institute of Technology, MIT, KAIST, Lulea University of Technology
  • CTU-CRAS-NORLAB (Score = 7): Czech Technological University, Université Laval
  • Coordinated Robotics (Score = 2): Coordinated Robotics, California State University Channel Islands, Oke Onwuka, Sequoia Middle School
  • Robotika (Score = 2): Robotika International, Robotika.cz, Czech University of Life Science, Centre for Field Robotics, Cogito Team

We congratulate all members of the team and we are proud of this incredible and historic achievement! Most importantly, we are excited to be part of this amazing community pushing the frontier of resilient robotic autonomy in extreme environments.



tags:


CERBERUS is a product of the collaboration between University of Nevada, Reno, ETH Zurich, Sierra Nevada Corporation, University of California, Berkeley, Flyability, the Norwegian University of Science and Technology (NTNU), and the University of Oxford.
CERBERUS is a product of the collaboration between University of Nevada, Reno, ETH Zurich, Sierra Nevada Corporation, University of California, Berkeley, Flyability, the Norwegian University of Science and Technology (NTNU), and the University of Oxford.

            AUAI is supported by:



Subscribe to Robohub newsletter on substack



Related posts :

Robot Talk Episode 166 – Evolving robot bodies and brains, with Rika Antonova

  16 Oct 2026
In the latest episode of the Robot Talk podcast, Claire chatted to Rika Antonova from the University of Cambridge about using artificial intelligence to develop complimentary robot hardware and algorithms.

Humanoid robots can sprint and box – but the real test is doing the laundry

  12 Oct 2026
No robot system can yet reliably handle the whole laundry process. Even folding remains a challenge.

Robot Talk Episode 165 – Robots behind the research, with Kushant Patel

  09 Oct 2026
In the latest episode of the Robot Talk podcast, Claire chatted to Kushant Patel from Brookhaven National Laboratory about robots that work in scientific facilities, such as particle accelerators.

What three days building a robot taught me

and   07 Oct 2026
The brief sounded simple: build a small autonomous robot, then make it survive increasingly difficult mazes. Simple... right? Until you actually have to build one.

Research could guide the future design of social robots

  05 Oct 2026
Giving robot companions animal-like or anime-style faces and voices could maximise their emotional appeal to users, according to new research which could help guide the growing field of social robotics.

Robot Talk Episode 164 – Accelerating robot learning, with Michelle Lu

  02 Oct 2026
In the latest episode of the Robot Talk podcast, Claire chatted to Michelle Lu from Vsim Technology about using simulation and artificial intelligence to teach robots new skills.

Robotics and automation? Informal reflections on familiar terms

IEEE experts reflect on what is meant by the term "robotics and automation".


↑


AUAI is supported by:







Subscribe to Robohub newsletter on substack




 















©2026.05 - Association for the Understanding of Artificial Intelligence