Robust Online Best-view Planning for Long-duration Underwater Autonomy

Abstract

The proposed research will develop and demonstrate new methods for simultaneous target detection, mapping, and localization for adva,nced long duration autonomy. The software will be delivered with a modular architecture developed in MATLAB and C++, demonstrated i,n Unreal Engine, and accompanied by clear and detailed technical documentation describing both the code and the numerical methods. C,ompeting algorithms will be developed and demonstrated in a common environment, thus providing multiple versions of each module alon,g with a detailed comparative and computational analysis. The algorithms with the best performance will be integrated with the modul,ar advanced autonomy architecture for maritime applications known as AVA. The PI and her students will work closely with Navy collab,orators and other performers in order to conduct AVA simulations and tests and, subsequently, transition the new software developed,in this project to maritime physical experiments. Various types of real andsynthetic sensor data, as well as physical and virtual te,st beds, will be used at every step of algorithm development, outlined below, to allow for incremental improvements to theory and nu,merical implementation driven by real maritime applications.

Document Details

Document Type
DoD Grant Award
Publication Date
Jul 13, 2022
Source ID
N000142212513

Entities

People

  • Silvia Ferrari

Organizations

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

Tags

Readers

  • Computational Modeling and Simulation
  • Robotics and Automation.
  • Unmanned Aerial System (UAS) Autonomous Capabilities and Mission Reconnaissance.