4 May 2014

Data Flood Helping the Navy Address the Rising Tide of Sensor Information

PDF file 3.6 MB 

Research Questions 
What changes across four dimensions — people, tools and technology, data and data architectures, and demand and demand management — will best position the Navy to solve its "big data" challenge and exploit big data's opportunities? 

Abstract

In the U.S. Navy, there is a growing demand for intelligence, surveillance, and reconnaissance (ISR) data, which help Navy commanders obtain situational awareness and help Navy vessels perform a host of mission-critical tasks. The amount of data generated by ISR sensors has, however, become overwhelming, and Navy analysts are struggling to keep pace with this data flood. Their challenges include extremely slow download times, workstations cluttered with applications, and stovepiped databases and networks — challenges that are only going to intensify as the Navy fields new and additional sensors in the coming years. Indeed, if the Navy does not change the way it collects, processes, exploits, and disseminates information, it will reach an ISR "tipping point" — the point at which its analysts are no longer able to complete a minimum number of exploitation tasks within given time constraints — as soon as 2016.

The authors explore options for solving the Navy's "big data" challenge, considering changes across four dimensions: people, tools and technology, data and data architectures, and demand and demand management. They recommend that the Navy pursue a cloud solution — a strategy similar to those adopted by Google, the Intelligence Community, and other large organizations grappling with big data's challenges and opportunities.


Key Findings

The Navy Faces Barriers to Making Sense of the Intelligence, Surveillance, and Reconnaissance Data Being Collected 

Challenges to the timely consumption of data include slow download times, shared communications pipelines, and large chunks of untagged raw data. 

Challenges to the accurate integration of data include stovepiped databases and networks, and cluttered analyst workstations. 

Dynamically Managing Analyst Workloads May Improve Analyst Productivity, but Only to a Certain Extent 

Today's tasking arrangements are, in the main, fixed and geographically based. Intelligence specialists in one location can become quickly overwhelmed with tasks. 

Tasking models in which tasks are automatically shared based on who is available to accept new tasking outperform today's model in terms of analyst productivity, but only in the short term. 
A Solution to the Navy's "Big Data" Challenge Must Involve Changes Across Four Dimensions: People, Tools and Technology, Data and Data Architectures, and Demand and Demand Management 

One potential solution involves adding more applications to analyst workstations, with the goal of helping analysts take advantage of the increased variety of data afforded by the proliferation of data types and databases. 

A second potential solution involves the physical consolidation of applications and their corresponding data and databases, with the goal of enabling a high level of interoperability. 

A third potential solution involves the virtual consolidation of databases, applications, widgets, services, and other elements into a cloud architecture, with the goal of limiting the transmission of raw data and of implementing a data strategy that includes the use of metadata. 

Recommendations

The Navy should pursue the third solution — a cloud strategy similar to those adopted by Google, the Intelligence Community, and other large organizations grappling with big data's challenges and opportunities. Specifically, the Navy should adopt the Intelligence Community's cloud approach, designing its next generation of intelligence, surveillance, and reconnaissance tools and systems to work with the National Security Agency's distributed cloud concept. 
Table of Contents 

Chapter One 
Big Data: Challenges and Opportunities 
Chapter Two 
What the Navy Wants from Big Data 
Chapter Three 
Barriers to Benefiting from Big Data 
Chapter Four 
Dynamically Managing Analyst Workloads 
Chapter Five 
Alternatives for Dealing with Big Data 
Chapter Six 
Analysis 
Chapter Seven 
Recommendations 
Appendix 
Additional Informat

No comments: