An Adaptive Inspection Planning Approach towards Routine Monitoring in Uncertain Environments
Vignesh Kottayam Viswanathan, Yifan Bai, Scott Fredriksson, Gajanan Sumeet Satpute, Christoforos Kanellakis, George Nikolakopoulos
AI summary
Problem
Pre-planned inspection routes frequently fail in dynamic environments due to discrepancies between historical maps and current conditions, such as new obstacles or evolving surfaces, which degrade data quality and require time-consuming re-mapping.
Approach
The method generates an initial global view plan from a historical map and continuously compares it with a reactive local path derived from real-time sensor data, triggering adaptive replanning only when significant deviations are detected.
Key results
- Successful real-world deployment on a quadrupedal robot in subterranean mines adapting to evolving surfaces and obstacles
- Maintained high viewpoint utility and desired viewing distances by switching plans based on a computed similarity threshold
- Outperformed non-adaptive baselines by rapidly correcting viewing offsets and improving inspection data quality
- Enabled fully onboard, real-time reactive adaptation without requiring complete environmental re-mapping
Why it matters
Enables reliable, high-quality autonomous visual inspection for mining and infrastructure monitoring in dynamic, unpredictable environments where static maps quickly become obsolete.
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 cov- erage objectives while enabling reactive adaptation to the local surface morphology. This enables the local autonomy to remain robust against environment uncertainty and complete the in- spection tasks. We validate the approach through deployments in real-world subterranean mines using quadrupedal robot. A supplementary media highlighting the proposed method can be found here https://youtu.be/6TxK8S_83Lw.