Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
97 tokens/sec
GPT-4o
53 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
5 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Mine Tunnel Exploration using Multiple Quadrupedal Robots (1909.09662v2)

Published 20 Sep 2019 in cs.RO

Abstract: Robotic exploration of underground environments is a particularly challenging problem due to communication, endurance, and traversability constraints which necessitate high degrees of autonomy and agility. These challenges are further exacerbated by the need to minimize human intervention for practical applications. While legged robots have the ability to traverse extremely challenging terrain, they also engender new challenges for planning, estimation, and control. In this work, we describe a fully autonomous system for multi-robot mine exploration and mapping using legged quadrupeds, as well as a distributed database mesh networking system for reporting data. In addition, we show results from the DARPA Subterranean Challenge (SubT) Tunnel Circuit demonstrating localization of artifacts after traversals of hundreds of meters. These experiments describe fully autonomous exploration of an unknown Global Navigation Satellite System (GNSS)-denied environment undertaken by legged robots.

User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (14)
  1. Ian D. Miller (15 papers)
  2. Fernando Cladera (16 papers)
  3. Anthony Cowley (3 papers)
  4. Shreyas S. Shivakumar (14 papers)
  5. Elijah S. Lee (10 papers)
  6. Laura Jarin-Lipschitz (5 papers)
  7. Akhilesh Bhat (2 papers)
  8. Neil Rodrigues (3 papers)
  9. Alex Zhou (9 papers)
  10. Avraham Cohen (3 papers)
  11. Adarsh Kulkarni (2 papers)
  12. James Laney (1 paper)
  13. Camillo Jose Taylor (11 papers)
  14. Vijay Kumar (191 papers)
Citations (73)

Summary

  • The paper introduces a novel decoupling of exploration from mapping, enabling effective path planning in GNSS-denied environments.
  • It details a coordinated multi-robot system using Ghost Robotics Vision 60 quadrupeds equipped with LiDAR and stereo cameras for real-time navigation and artifact detection.
  • Field experiments validate the system's robustness while highlighting challenges in sensor reliability and motion planning on complex terrains.

Mine Tunnel Exploration using Multiple Quadrupedal Robots

The research paper titled "Mine Tunnel Exploration using Multiple Quadrupedal Robots" provides a detailed examination of deploying legged robots for autonomous exploration of subterranean environments. The challenges intrinsic to underground exploration, such as communication limitations and the demand for high degrees of autonomy, are addressed through a sophisticated system design involving a coordinated team of quadrupedal robots. This paper delineates advances in multi-robot exploration and demonstrates experimental results from the DARPA Subterranean Challenge (SubT) Tunnel Circuit.

Summary of Contributions and Methodology

The paper's contributions are concentrated in several areas: exploration and planning algorithms, communication systems, and detection mechanisms. A notable aspect of this paper is the focus on decoupling exploration from mapping, which enables a robust approach to path planning in GNSS-denied environments. The exploration algorithm is predicated on real-time tunnel detection, leveraging instantaneous depth panoramas and tracking these features via Extended Kalman Filters (EKFs) to make autonomous navigation decisions.

The multi-robot system employs Ghost Robotics Vision 60 quadrupeds equipped with diverse sensors, including LiDAR for mapping and stereo cameras for pose estimation and object detection. This sensor suite, coupled with on-board computation, enables real-time navigation and artifact detection in environments that encompass challenging terrains, such as concrete passages and railway ties. The software architecture is built on the Robot Operating System (ROS), facilitating modular subsystem integration and communication resilience in low-bandwidth conditions.

The paper highlights a distributed database approach for data sharing among robots, essential for ensuring data consistency in environments with intermittent connectivity. Such decentralized communication allows individual robots to operate with high autonomy while sharing critical insights with other nodes in the mesh network. This system is complemented by user interfaces that furnish operators with real-time information on robot status and detected artifacts, facilitating decision-making during exploration missions.

Experimental Results and Observations

Field experiments carried out in the National Institute for Occupational Safety and Health (NIOSH) Experimental Mine provided empirical validation for the system's capabilities. In laboratory settings, the robot teams demonstrated successful navigation and artifact detection with low localization error. However, challenges such as sensor failures and hazardous terrain inhibited some missions, emphasizing areas for improvement in robot reliability and planning robustness.

In practice, robots periodically fell due to the planner's overconfidence in its traversal cost estimates, particularly on complex surfaces like gravel. This issue underlines the necessity for enhancing motion planning algorithms, potentially by incorporating vision-based terrain analysis and motion-primitive planning. Despite these hurdles, the robots accomplished significant traversal distances, sometimes as far as several hundred meters, underscoring the potential of legged robotics in subterranean applications.

Implications and Future Research Directions

The paper advances understanding in autonomous exploration using legged robots, contributing practically to technology development for subterranean exploration. The presented system offers a foundational framework adaptable to other exploration scenarios beyond mining, leveraging multi-robot collaboration and decentralized communication.

Future research directions recommended by the authors include refining terrain analysis to improve robustness on diverse surfaces, integrating optical data into the mapping pipeline, and exploring collaborative strategies for multiple robots to optimize exploration coverage. These avenues could bolster system performance and broaden the applicability of quadrupedal robotics in unmanned exploration missions.

Overall, this work serves as a comprehensive resource for robotics researchers focusing on autonomous exploration, presenting both a step forward in addressing complex subterranean challenges and a call for further enhancement of autonomous systems in GNSS-denied environments.

Youtube Logo Streamline Icon: https://streamlinehq.com