Coordination of a Multi-Platform UUV/ASV System: Low-Cost Experimental and Simulation Test Environment with Fuzzy Logic Based (AI) Autonomy Evaluation
Abstract
Coordination of a Multi-Platform UUV/ASV System: Low-Cost Experimental and Simulation Test Environment with Fuzzy Logic Based (AI) Autonomy Evaluation The development of autonomous marine vehicles, particularly the coordination of autonomous UUVs, continues to be in high demand. One of the significant challenges in the development of autonomous marine vehicles is developing the ability of the vehicle to perceive its environment for executing its mission, including self-localization, obstacle avoidance, area/mapping, particularly for the purposes of obstacle/collision avoidance. For this purpose, the PI proposes to develop a low-cost test bed environment with three research goals: (1) to develop a simulation testing environment to analytically observe and predict UUV perception capabilities and its performance on overall system autonomy; (2) to further develop an experimental multi-vehicle test platform (for laboratory and field testing) in which (2A) UUV perception and autonomy capabilities can be physically tested and (2B) a low-cost sensor system will be designed and implemented to enable autonomous coordination of multiple UUVs; and (3) to develop an Artificial Intelligence (AI) based metric system (via Fuzzy Logic and Fuzzy Set Theory) to numerically evaluate vehicle autonomy performance. The proposed simulation test environment will be developed to test and compare autonomy algorithms based upon user-defined mission criteria and canonical autonomy subtask capabilities. Previous automated path planning techniques (from previous NEEC grant results) will be further developed and implemented. Fault Detection and Isolation/Mitigation (FDI/M) techniques will developed (with the aid of redundant sensor banks) to not only improve vehicle perception capabilities, but to also improve the quality of sensor measurements needed for reliable automatic feedback-based vehicle control. In addition to the technical developments and contributions of this research, it is expected that this project will require the active participation of 3 generations of approximately 24 interdisciplinary (EE/CE, CS, ME, OE) undergraduate and graduate students annually (all together about 66 students).
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 13, 2025
- Source ID
- N001742010006
Entities
People
- May-win Thein
Organizations
- United States Navy
- University of New Hampshire