An Adaptive Inspection Planning Approach Towards Routine Monitoring in Uncertain Environments

TL;DR

The paper tackles robust autonomous visual inspection in uncertain subterranean environments where historical maps ($\mathbb{M}_H$) may diverge from current surfaces ($\mathbb{M}_C$) due to obstructions and evolving morphology ($\mathbb{M}^{\delta}$). It proposes a hierarchical framework that fuses global view planning on $\mathbb{M}_H$ with a local reactive planner using real-time sensor data to preserve viewing constraints. A Fréchet-distance based similarity metric $\Gamma_s$ between the global visiting plan $\Pi^k_{GVP}$ and the local plan $\Pi^k_{LVP}$ triggers a Kabsch alignment yielding $\hat{\Pi}_{GVP}$, which is tracked by a nonlinear MPC; the approach is demonstrated on a quadrupedal robot in subterranean mines and validated in simulations. Validated in subterranean mines and simulations, the approach demonstrates improved viewpoint utility and adaptive behavior over non-adaptive baselines, enabling robust inspection in dynamically changing environments.

Abstract

In this work, we present a hierarchical framework designed to support robotic inspection under environment uncertainty. By leveraging a known environment model, existing methods plan and safely track inspection routes to visit points of interest. However, discrepancies between the model and actual site conditions, caused by either natural or human activities, can alter the surface morphology or introduce path obstructions. To address this challenge, the proposed framework divides the inspection task into: (a) generating the initial global view-plan for region of interests based on a historical map and (b) local view replanning to adapt to the current morphology of the inspection scene. The proposed hierarchy preserves global coverage objectives while enabling reactive adaptation to the local surface morphology. This enables the local autonomy to remain robust against environment uncertainty and complete the inspection tasks. We validate the approach through deployments in real-world subterranean mines using quadrupedal robot.

Figures (12)

• Figure 1: Examples of robotic inspection application in subterranean environments.
• Figure 2: Framework of the proposed methodology.
• Figure 3: A visual representation of the initial view plan generation scheme for a defined region of interest.
• Figure 4: A visual schema of the proposed adaptive inspection planning approach in uncertain environments.
• Figure 5: A visualization of the subterranean environment.