NICOP - Intent detection algorithm development for use on resource limited devices for Maritime Domain Awareness
Abstract
Intent detection algorithm development for use on resource limited devices for Maritime Domain Awareness:a. Data exfiltration from remote sensors faces challenges from such issues as bandwidth, power, and connectivity windows. This is an issue that is confronted both in operations and in pure research environments. With the miniaturization of computers and improvements in processing speeds, there is now the possibility of improved pre-processing, so that sensors only send back the most critical data. This project focuses on this issue via the lens of maritime traffic data, but it is something with application to oceanographic data collection as well. The team will also be using open hardware boards (fabricated in Brazil), which offer even more options for customizing pre-processing solutions. b. The US navy faces two seemingly contradictory challenges with maritime surveillance data- in some locations the data is so dense, that it challenges our ability to process it and find interesting targets (particularly with video data). Alternatively, large parts of the globe are effectively blank because of the logistical challenges of keeping sensors operating. This project looks at the core issues around sensor design and data processing, with an eye to developing solutions that can be deployed in these "blank spot" areas. c. NRL 8000 (Jay Middour group) d. Improved collection of maritime data in empty spaces.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 23, 2016
- Source ID
- N629091612183
Entities
People
- Waldo Kleynhans
Organizations
- Office of Naval Research
- United States Navy