Ethologically inspired reactive exploration of coral reefs with collision avoidance: Bridging the gap between human and robot spatial understanding of unstructured environments

摘要

We present a behavioral approach for autonomous robotic exploration of marine habitat with collision avoidance given little or no prior information. In our previous work, a vision-based reactive navigation paradigm with a predefined forward direction allowed an underwater robot to avoid unexpected obstacles. In this work, we have now incorporated visual perceptive invariants to guide the navigation as a human diver would do, thus eliminating the need for a predefined trajectory or exact pose estimation, i.e., we are adding behavioral patterns to turn the navigation into an exploration. Our exploration architecture consists of a tracker of regions of interest and, a collision avoidance behavior. During the exploration, the robot's navigation system is guided towards the most relevant region while continuously checking for free space (water) to navigate. The proposed framework successfully combines both behaviors despite their opposite nature (moving towards a region to explore it while avoiding collisions). We have implemented our approach on an environment-friendly robot, which uses fins instead of propellers to allow for a non-invasive and cautious exploration. Results of sea trials performed at different locations and depths demonstrate the feasibility of our approach.