Katie Toth (NIWC-LANT Data Scientist), Alex Wu (UCSD Data Science Graduate Student), Eric Sen (UCSD Data Science Graduate Student), Dr. Henrik Christensen (UCSD Professor " Computer Science and Engineering, Director " Contextual Robotics Institute), David Byres (NAVWAR Data Science Team Project Manager), Alex Hancock (UCSD Data Science Graduate Student) Not pictured: Dr. Ilkay Altintas (UCSD Data Science Institute Professor)
NAVWAR partners with UCSD, explores data science for improved decision making
by Elisha Gamboa,
Naval Information Warfare Systems Command Public Affairs
SAN DIEGO (NNS) -- Naval Information Warfare Systems Command (NAVWAR) is collaborating with the University of California, San Diego (UCSD) to study the use of data science techniques and analysis to increase competitive intelligence and improve decision-making Navy-wide.
Established in mid-2018, the NAVWAR data science team comprises NAVWAR’s Logistics and Fleet Readiness Competency, UCSD’s Halcolu Data Science Institute, and multiple Navy organizations focused on examining data-driven decision-making to produce calculated insights and solve complex problems for increased operational efficiency.
“Modern advancements in technology have drastically increased the amount of information that is being collected and stored,” said NAVWAR Executive Director Pat Sullivan. “That information has the potential to enable a competitive edge for the fleet, but only if the Navy has the ability to interpret it and apply it to the current environment.”
The NAVWAR data science team is made up of experts from the following institutes and organizations:
- NAVWAR Logistics and Fleet Readiness Competency
- UCSD Halcolu Data Science Institute
- Naval Information Warfare Center Pacific (NIWC Pacific)
- Naval Information Warfare Center Atlantic (NIWC Atlantic)
- Program Executive Office for Command, Control, Communications, Computers and Intelligence (PEO C4I)
- Naval Surface Warfare Center (NSWC) Corona
- Southwest Regional Maintenance Center (SWRMC)
- Naval Surface Forces, US Pacific Fleet (SURFPAC)
- Naval Air Forces, US Pacific Fleet(AIRPAC)
- Naval Education and Training Command (NETC)
UCSD established the Halcolu Data Science Institute in 2018, an academic unit focused on training students, faculty and industrial partners to use data science in new and novel ways that will allow them to understand some of the world’s most pressing problems.
“Partnering with UCSD has greatly accelerated our efforts to become more innovative in our approach to answering difficult questions through data,” said David Byres, NAVWAR information technology (IT) specialist and project manager for the data science team.
The NAVWAR data science team established a project team and developed basic processes for team collaboration, data extraction, exploration and analysis.
Together they built on a hypothesis that one could combine data, in the form of problem descriptions, from the Regional Maintenance Center’s support database with other incident management data to reveal trends and produce insights about possible solutions for new fleet support incidents.
“During their research, the team applied natural language processing techniques to relate short problem description narratives among different and unrelated fleet support IT systems,” said Byres. “By relating these separate sources of information, potential solutions to fleet problems that are ‘bottled up’ in our IT systems could potentially be leveraged to more rapidly identify solutions to new problems.”
The project team concluded this effort in June of 2019.
Additionally, the team has taken on a more involved data science endeavor. Expanding on a RAND Corporation study titled “Navy Network Dependability: Models, Networks and Tools” and applying it to current fleet challenges, the team is analyzing factors that influence the “user perceived dependability” of complex NAVWAR systems.
While network dependability is determined by the availability and reliability of IT systems, user perceived dependability refers to how dependable a user understands or perceives a system to be.
Today, the Navy is increasingly dependent on networks and associated net-centric operations to conduct its missions. To increase their understanding of network dependability and user perceived dependability across ship and multi-ship networks, the NAVWAR data science team is investigating the relationship among existing sources of data including hardware, software and human factors.
More specifically, the team is trying to understand which fleet IT staffing and training factors most affects a ship’s ability to support their Consolidated Afloat Network and Enterprise Services (CANES). In other words, what makes some ships more successful in operating CANES than others?
“We’re working closely with our many project partners to arrive at a much deeper understanding of what factors most impact a crew’s ability to operate and maintain their complex shipboard IT networks,” said Byres. “We’ve really just scratched the surface in terms of identifying and working with Navy data that might provide answers to our most basic questions.”
The long-term project is currently ongoing, with an end date projected for fiscal year 2020.
As the demand for data science in defense continues to grow, NAVWAR will continue to pursue the use of more advanced data science techniques and analytics to better capitalize on existing information to ensure the Navy can fight and win today and in the coming decades.
NAVWAR identifies, develops, delivers and sustains information warfighting capabilities and services that enable naval, joint, coalition and other national missions operating in warfighting domains from seabed to space. NAVWAR consists of more than 11,000 active duty military and civil service professionals located around the world.
Military News | Navy News | NAVWAR partners with UCSD