Information-driven Guidance & Control of Hetergenous Underwater Sensor Networks for Adaptive Target Detection & Classification

Abstract

The proposed research will develop and demonstrate an information-driven approach for the control of sonar-equipped unmanned underwater vehicles (UUVs) that are deployed for the purpose of detecting, classifying, reacquiring, and identifying unexploded ordnances (UXO) subject to environmental variability and uncertainty. Rapid environmental assessment (REA), surveying, and target detection, classification, and reacquisition missions typically involve multiple sensing assets whose performance is greatly influenced by the local environmental conditions and by the position and orientation of the UUV relative to the target and the seafloor. At the same time, these missions require safe underwater navigation and exploitation of the underwater topography for accurate geo-referencing, and of four-dimensional geophysical fluid structures for minimizing the energy needed for propulsion. For example, bathymetric data can be used for tactical navigation, as well as for determining the optimal positioning and orientation of the active sonar to circumvent reduced performance due to bottom clutters and underwater terrain shadows. In many cases, the risk of collisions and energy consumption must be minimized while sensing objectives, such as bottom surveying and target classification and identification, must be simultaneously maximized. The proposed research will develop novel information-driven navigation algorithms based on nonparametric Bayesian models and computational geometry to improve overall tactical navigation and mission planning, while enabling longer and safer missions by minimizing energy consumption and risk of collisions. The outcome of the proposed research activity will be a suite of vehicle control algorithms applicable to open-ocean seagliders (e.g. SLOCUM) or costal operation UUVs (e.g. REMUS) characterized by different sensing modalities, such as side-scan and forward-looking sonar. The Nodes developed in this research will be designed to provide heading/speed/depth, waypoint homing, line following, and survey area command inputs to MCMES vehicle control default services, based on inputs received from the MCMES voxel discretization of the environment.

Document Details

Document Type
DoD Grant Award
Publication Date
Aug 08, 2016
Source ID
N000141512595

Entities

People

  • Silvia Ferrari

Organizations

  • Cornell University
  • Office of Naval Research
  • United States Navy

Tags

Readers

  • Acoustical Oceanography.
  • Sensor Fusion and Tracking Systems.
  • Unmanned Aerial System (UAS) Autonomous Capabilities and Mission Reconnaissance.

Technology Areas

  • AI & ML
  • AI & ML - Autonomous Systems
  • Autonomy