NeBula: Quest for Robotic Autonomy in Challenging Environments; TEAM CoSTAR at the DARPA Subterranean Challenge
Abstract
This paper presents and discusses algorithms, hardware, and software architecture developed by the TEAM CoSTAR (Collaborative SubTerranean Autonomous Robots), competing in the DARPA Subterranean Challenge. Specifically, it presents the techniques utilized within the Tunnel (2019) and Urban (2020) competitions, where CoSTAR achieved 2nd and 1st place, respectively. We also discuss CoSTAR's demonstrations in Martian-analog surface and subsurface (lava tubes) exploration. The paper introduces our autonomy solution, referred to as NeBula (Networked Belief-aware Perceptual Autonomy). NeBula is an uncertainty-aware framework that aims at enabling resilient and modular autonomy solutions by performing reasoning and decision making in the belief space (space of probability distributions over the robot and world states). We discuss various components of the NeBula framework, including: (i) geometric and semantic environment mapping; (ii) a multi-modal positioning system; (iii) traversability analysis and local planning; (iv) global motion planning and exploration behavior; (i) risk-aware mission planning; (vi) networking and decentralized reasoning; and (vii) learning-enabled adaptation. We discuss the performance of NeBula on several robot types (e.g. wheeled, legged, flying), in various environments. We discuss the specific results and lessons learned from fielding this solution in the challenging courses of the DARPA Subterranean Challenge competition.
Keywords
Cite
@article{arxiv.2103.11470,
title = {NeBula: Quest for Robotic Autonomy in Challenging Environments; TEAM CoSTAR at the DARPA Subterranean Challenge},
author = {Ali Agha and Kyohei Otsu and Benjamin Morrell and David D. Fan and Rohan Thakker and Angel Santamaria-Navarro and Sung-Kyun Kim and Amanda Bouman and Xianmei Lei and Jeffrey Edlund and Muhammad Fadhil Ginting and Kamak Ebadi and Matthew Anderson and Torkom Pailevanian and Edward Terry and Michael Wolf and Andrea Tagliabue and Tiago Stegun Vaquero and Matteo Palieri and Scott Tepsuporn and Yun Chang and Arash Kalantari and Fernando Chavez and Brett Lopez and Nobuhiro Funabiki and Gregory Miles and Thomas Touma and Alessandro Buscicchio and Jesus Tordesillas and Nikhilesh Alatur and Jeremy Nash and William Walsh and Sunggoo Jung and Hanseob Lee and Christoforos Kanellakis and John Mayo and Scott Harper and Marcel Kaufmann and Anushri Dixit and Gustavo Correa and Carlyn Lee and Jay Gao and Gene Merewether and Jairo Maldonado-Contreras and Gautam Salhotra and Maira Saboia Da Silva and Benjamin Ramtoula and Yuki Kubo and Seyed Fakoorian and Alexander Hatteland and Taeyeon Kim and Tara Bartlett and Alex Stephens and Leon Kim and Chuck Bergh and Eric Heiden and Thomas Lew and Abhishek Cauligi and Tristan Heywood and Andrew Kramer and Henry A. Leopold and Chris Choi and Shreyansh Daftry and Olivier Toupet and Inhwan Wee and Abhishek Thakur and Micah Feras and Giovanni Beltrame and George Nikolakopoulos and David Shim and Luca Carlone and Joel Burdick},
journal= {arXiv preprint arXiv:2103.11470},
year = {2021}
}
Comments
For team website, see https://costar.jpl.nasa.gov/. Accepted for publication in the Journal of Field Robotics, 2021