West Virginia University
Enhancing Robotic Exploration through Semantically-Guided Sampling Strategies
Abstract
dc:description.abstractFrom space and deep-sea exploration to disaster response and environmental monitoring, autonomous robots are essential for advancing science, improving safety, and addressing critical challenges. This dissertation introduces a novel open-source strategy for autonomous robotic exploration: the Semantically-Guided Exploration (SGE) framework. Designed for ground vehicles, SGE integrates semantic understanding into the autonomous exploration process, improving decision-making in complex environments. Specifically, the proposed sampling-based approach uses the information from the semantic segmentation of RGB images and depth images to guide the robot's selection of exploration goals. This method enables the robot to steer away from potential dangers such as large rocks and water, while prioritizing a specific type of terrain or objective, such as staying on trails, making exploration safer and more flexible. Additionally, an exploration manager framework is proposed to process these waypoints. It optimizes viewpoint selection via a Traveling Salesman Problem (TSP) formulation in a receding-horizon manner, ensuring robots make intelligent, context-driven decisions while navigating uncertain terrains. Furthermore, a new learning-based direct sampling method is presented, which aims to mimic human-like exploration behaviors, by teaching robots to select waypoints from human-provided examples directly from raw data inputs. The methods are extensively evaluated in both in simulation and in real-world settings, including the university campus, indoor corridors, and underground mining environments. Experimental results validate the framework's effectiveness, demonstrating that SGE provides significant advantages increasing flexibility and safety, while maintaining competitive performance with state-of-the-art exploration techniques in benchmarked tests evaluating volumetric data and exploration distance. Overall, the SGE framework enhanced environmental understanding and decision-making capabilities, enabling autonomous robots to operate effectively in less constrained and more challenging environments.
Degree
thesis:*- Name thesis:degree_name
- PhD
- Level thesis:degree_level
- Dissertation
- Discipline thesis:degree_discipline
- Mechanical and Aerospace Engineering
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Arend Tatsch, Christopher Alexander
- Contributors dc:contributor
-
- Yu Gu
- Jason Gross
- Guilherme Pereira
- Gianfranco Doretto
- Xi Yu
Subjects
dc:subject × 9Identifiers
dc:identifier.*- Identifier
-
<p><a href="https://orcid.org/0000-0002-9817-1128">https://orcid.org/0000-0002-9817-1128</a></p>
https://researchrepository.wvu.edu/etd/12637