Explore Locally, Plan Globally: A Path Planning Framework for Autonomous Robotic Exploration in Subterranean Environments

This paper presents a path planning strategy for the autonomous exploration of subterranean environments. Tailored to the specific challenges and particularities of underground settings, and especially the fact that they often are extremely large in scale, tunnel-like, narrow and multi-branched, the proposed method employs a bifurcated local- and global-planning design to enable exploration efficiency and path planning solution resourcefulness. The local planner builds on top of minimum-length random trees and efficiently identifies collision-free paths that optimize for exploration within a local subspace, while simultaneously ensuring enhanced obstacle clearance and thus safety. Accounting for the robot endurance limitations and the possibility that the local planner reaches a dead-end (e.g. a mine heading), the global planner utilizes an incrementally built graph to search within the full range of explored space and is engaged when the robot should be repositioned towards a frontier of the exploration space or when a return-to-home path must be derived. The proposed approach is field evaluated in a set of deployments in an exploratory underground mine drift in Northern Nevada.