Intent Recognition Software for Naval Domains

Case ID:
NVRIC20-005

Background

Current vessel monitoring systems include Vessel traffic management systems (VTMS) and vessel traffic monitoring information systems (VTMIS). Both have been around for many years however, due to the increased complexities in the naval domain, enhanced systems are needed in order to maintain the safety of navy personnel and assets. According to the National Strategy for Maritime Security, developing new systems and processes to attain both an understanding of maritime activity as well as maintain safety are key cornerstones of the organization. Early classification of hostile behavior in maritime domain is something that is currently being heavily focused on as an area of improvement and can allow an action to be taken in a timely and efficient manner. Hidden Markov Models are emerging as a beneficial tool because of its capabilities of detecting multiple hostile intents during the early stages of the behaviors, giving ample time for action to be taken.

Description

Researchers at the University of Nevada, Reno have developed software that provides the ability to detect various types of intentions (which may pose threats or could be harmless) for boats, from data gathered from on-board sensory systems. 

Advantages

  1. Ability to detect various types of intentions

  2. Autonomously detects potential threats posed by other ships/boats

  3. The prediction system can be integrated with existing technology

 

Patent Information:
For Information, Contact:
Cara Baird
Licensing Associate
University of Nevada, Reno
carab@unr.edu
Inventors:
Monica Nicolescu
Mircea Nicolescu
Sushil Louis
Keywords:
Software