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.

Autonomous Teamed Exploration of Subterranean Environments using Legged and Aerial Robots

“…In response, groups around the world have proposed novel methods both for single-and multi-robot exploration. This includes techniques for quadruped robots [28], methods tailored to fast exploration using aerial platforms [27], schemes for both legged and flying systems [21], hierarchical frameworks to exploit dense local and sparse global information [45], multi-robot exploration strategies [34], approaches exploiting underground mine and cave topologies [46,47], and significant field robotics work [48][49][50][51]. Motivated by the importance of autonomous exploration and tailored to the teamed deployment of legged and flying robots inside subterranean settings, this work contributes two methods, namely on teamed exploration coordination and single-robot planning that enable resilient multi-robot teaming and reliable single-robot operation when communication to and from a robot is not possible, ability to negotiate challenging terrain and capacity to map diverse and large-scale geometries.…”

Direct LiDAR Odometry: Fast Localization with Dense Point Clouds

“…Our system also optionally accepts a relative prior from IMU in a loosely-coupled fashion to further improve accuracy in the optimization process, which can help especially during aggresive rotational motions. The reliability of our approach is demonstrated through extensive tests on multiple computationally-limited robotic platforms in several environments as part of Team CoSTAR's research and development efforts for the DARPA Subterranean Challenge in support of NASA JPL's Networked Belief-aware Perceptual Autonomy (NeBula) framework [12], in which DLO was the primary state estimation component on our fleet of autonomous aerial vehicles (Fig. 1A).…”

Learning Risk-aware Costmaps for Traversability in Challenging Environments

“…observations, or processed map data (e.g. represented as m = (m (1) , m (2) , • • • ) where m i is the i-th element of the map). In the case of a 2D or 2.5D representation, i represents the cell index.…”

“…A new data sample was added to the dataset at approximately 0.5Hz, while the average top speed of the robot is 1m/s. Data was collected using Boston Dynamics's Spot legged robot equipped with JPL's NeBula payload [1]. The payload includes onboard computing, one VLP-16 Velodyne LiDAR sensor, and a range of other cameras and sensors.…”