Quest for Robotic Autonomy in Extreme Environments
The Search for Life
Is (or was) there life beyond Earth? The answer to this question leads underground on planetary bodies in our solar system. Planetary subsurface voids are one of the most likely places to find signs of life (both extinct and extant). Subsurface voids are also one of the main candidates for future human habitat for colonization beyond Earth. To this end, TEAM CoSTAR is participating in the DARPA Subterranean Challenge to develop fully autonomous systems to explore subsurface voids with a dual focus on planetary exploration and terrestrial applications in search and rescue, mining industry, and AI/Autonomy in extreme environments.
The DARPA Subterranean or “SubT” Challenge is a robotic competition that seeks novel approaches to rapidly map, navigate, and search underground environments. The competition spans a period of three years. CoSTAR is a DARPA-funded team participating in the systems track developing and implementing physical systems that will be tasked with the traversal, mapping, and search in various subterranean environments: including natural caves, mines, and urban underground.
TEAM CoSTAR [Collaborative SubTerranean Autonomous Resilient Robots] is a collaboration between NASA’s JPL, MIT, Caltech, KAIST, LTU, and several industry partners (see below). TEAM CoSTAR with more than 60 key members aims at revolutionizing how we operate in the underground domains and subsurface voids for both terrestrial and planetary applications.
NeBula Autonomy Solution
To address various technical challenges across multiple domains in autonomous exploration of extreme environment, CoSTAR develops a unified modular software system, called NeBula (Networked Belief-aware Perceptual Autonomy). NeBula is specifically designed to address stochasticity and uncertainty in various elements of the mission, including sensing, environment, motion, system health, communication, among others. NeBula has been implemented on multiple heterogeneous robotic platforms (wheeled, legged, tracked and flying vehicles).
- Verifiable autonomy: NeBula develops an autonomy architecture that translates the mission specifications into single- or multi-robot behaviors using verifiable formal methods. NeBula quantifies risk and trust in this process by taking uncertainty in robot motion, control, sensing, and environment into account when abstracting activities and behaviors.
- Modularity and mobility-based adaptation: NeBula focuses on a modular design to enable adaptation to various mobility platforms (legged, flying, wheeled, and tracked) and various computational capacities.
- Resilient Navigation: NeBula develops a resilient navigation system by degeneracy-aware fusions of various complementary sensing modalities, including vision, IMU, lidar, radar, contact sensors, and ranging systems (e.g., magneto-quasi static signals and UWBs). The system can switch between modalities based on the availability and the information content in various channels.
- Single- and multi-robot SLAM and dense mapping: NeBula develops large-scale SLAM solvers and 3D mapping frameworks using confidence-rich mapping methods to provide precise topological and geometrical maps of the subsurface caves and mine networks.
- Disruption tolerant mesh communication: NeBula extends disruption tolerant communication (DTN) to enable a self-healing coordination of a robotic swarm with lightweight stationary (deployable) and mobile communication nodes.
- Autonomous skill learning: NeBula applies and extends reinforcement learning and in general machine learning methods to enable fast and safe robot motions in perceptually-degraded environments.
Space Application of NeBula
As part of NASA’s B R A I L L E project, CoSTAR robots search for life and mineral resources in North America’s largest network of volcanic caves. These activities will help NASA prepare for future life-detection missions to caves on Mars and other planetary bodies.
For related details of NeBula deployment in support of R&D efforts and space missions targeting our moon, Europa, Enceladus, and other bodies, see here.
Take Part, join us
Interested in joining the team? Please contact us at email@example.com for more details on how to sponsor and support the team.
Technical Lead – Principal Investigator
- Dr. Ali-akbar Agha-mohammadi
- NASA’s Jet Propulsion Laboratory, Caltech
Charles C. Gates Jr.–Franklin Thomas Laboratory, Caltech
- Prof. Joel W. Burdick
- California Institute of Technology
- Laboratory for Robotics and Bioengineering
Laboratory for Information & Decision Systems, MIT
- Prof. Luca Carlone
- Massachusetts Institute of Technology
- Laboratory for Information & Decision Systems
Unmanned Systems Research Group, KAIST
- Prof. "David" Hyunchul Shim
- Korea Advanced Institute of Science and Technology
- Unmanned Systems Research Group
Robotics Team, LTU
- Prof. George Nikolakopoulos
- Luleå University of Technology
Below are several papers describing recent results from the NeBula framework:
- "Deep Learning Tubes for Tube MPC," Robotics: Science and Systems (RSS), 2020. [Link]
- “Confidence-rich 3D Grid Mapping,” International Journal of Robotics Research (IJRR), vol.38, pp.1352-1374, 2019. [Link]
- "Towards Resilient Autonomous Navigation of Drones,” The International Symposium on Robotics Research (ISRR). Hanoi, Vietnam, October, 2019.
- “Bi-directional Value Learning for Risk-aware Planning Under Uncertainty,” IEEE Robotics and Automation Letters (RA-L), vol.4, no.3, pp.2493-2500, March, 2019. [Link]
- "LAMP: Large-Scale Autonomous Mapping and Positioning for Exploration of Perceptually-Degraded Subterranean Environments," IEEE International Conference on Robotics and Automation (ICRA), 2020. [Link]
- “Contact Inertial Odometry: Collisions are your Friend,” The International Symposium on Robotics Research (ISRR), Hanoi, Vietnam, October, 2019. [Link]
For a full list of NeBula publications, please click here.