| |
| |
Research Projects >> Intelligent Controls and Communications Systems >> Current Research Project >>
Cooperative Autonomous Underwater Vehicles used to Search Large Ocean Areas for Mines | | Status:
Current Research Project
Research Category:
Intelligent Controls and Communications Systems
Research Center:
Microelectronics Research and Communications Institute
Sponsor(s)
Office of Naval Research (ONR)
Primary Researchers:
Dean B. Edwards,
James F. Frenzel,
Michael O'Rourke,
Michael J. Anderson and
Richard W. Wall.
Research Engineer:
Thomas A. Bean.
Research Associates:
John Canning and
Eric Hake.
Graduate Student Research Assistants:
Hans Leidenfrost,
Benjamin Armstrong,
Nicodemus Hallin,
Daniel Meyer,
Jesse Pentzer,
Patrick Ray and
Clayton Tye.
Undergraduate Student Research Assistants:
David Billin,
Benjamin Johnson and
Mark Kinney.
Duration:
February 1, 2008 to July 31, 2009 |
The overall objective of the UC2I Program is to find all mines in a large area of the ocean in an efficient and expedient manner via a collaborative, adaptive strategy for identifying and assessing mine-like objects (MLOs) that enables unmanned underwater vehicles (UUVs), which we also mean to be autonomous, to share resources in order to reduce search time and more evenly distribute energy consumption. To this end, we are developing advanced collaborative behaviors needed by UUVs conducting mine countermeasures, including (1) multi-vehicle sensing of MLOs; (2) a supervisory planner for mission optimization; and (3) language-based communication. This requires developing collaborative behaviors and related language with the CAD/CAC systems for identifying MLOs. The supervisory planner uses fuzzy logic and a rule base to allocate resources while managing mission time and vehicle energies, and is initially trained through computer simulations. We are also investigating multi-objective optimization functions for use in the planner. Because all vehicles use the same logic for language interpretation, this logic can also be used in the planner to help develop scenarios and estimate future vehicle energies and total mission time. We call this “language centered intelligence” and we plan to use this approach to simplify the planner logic. Natural language-based agent communication languages will be used as a medium for overcoming message error. Messages that contain errors often also contain good information that could be exploited to correct the error and improve communication reliability.
The goal of our research is to develop and apply autonomous collaboration among UUVs to search for underwater mines. Formation-flying and underwater communication among multiple vehicles will allow automated deployment, coordinated sensing, diversion of vehicles to obtain high-resolution information, and mapping to optimize return of information given expenditures of time and energy.
| |
